pipeline corrosion and coating failure under alternating

University of Calgary

PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2016

Pipeline Corrosion and Coating Failure under

Alternating Current Interference

Kuang, Da

Kuang, D. (2016). Pipeline Corrosion and Coating Failure under Alternating Current Interference

(Unpublished doctoral thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/27361

http://hdl.handle.net/11023/2968

doctoral thesis

University of Calgary graduate students retain copyright ownership and moral rights for their

thesis. You may use this material in any way that is permitted by the Copyright Act or through

licensing that has been assigned to the document. For uses that are not allowable under

copyright legislation or licensing, you are required to seek permission.

Downloaded from PRISM: https://prism.ucalgary.ca

UNIVERSITY OF CALGARY

Pipeline Corrosion and Coating Failure under Alternating Current Interference

by

Da Kuang

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

GRADUATE PROGRAM IN MECHANICAL AND MANUFACTURING ENGINEERING

CALGARY, ALBERTA

APRIL, 2016

© Da Kuang 2016

ii

Abstract

The results demonstrate that the size of the defect is critical to AC corrosion of steel occurring

at the defect base. Generally, circular and triangular defects are associated with the most and least

negative direct current (DC) potentials as well as the largest and smallest anodic current densities,

respectively. Local pits can be initiated on the steel at sufficiently high AC current densities in

both high pH and neutral pH solutions. Mechanistic models are proposed to illustrate the pitting

initiation in the presence of AC, where AC can affect passive film or corrosion product layer

formed on the steel.

Furthermore, the effect of AC on the properties and performance of epoxy coatings was studied

in a simulated soil solution and a conceptual model was developed to illustrate the mechanistic

aspect of AC induced coating degradation. The applied CP can be shielded by coating disbondment

through monitoring the local potential and solution pH under disbonded coating on a steel.

Then the CP shielding behavior of pipeline coatings, i.e., high density polyethylene (HDPE)

and FBE, was investigated. The structure of HDPE does not change upon CP permeating test.

Conversely, there are obvious changes of the functional groups in FBE. Water uptakes into the

coating occur continuously with time. The CP permeation through FBE coating is time dependent.

Moreover, the effect of AC interference on the CP permeation into the disbonding crevice was

investigated. At small AC current densities, AC results in an enhancement of permeation of CP

current into the crevice. However, with the increase of AC current density, corrosion product

generates and deposits in the solution, blocking the ionic diffusion and CP permeation.

Finally, the effects of AC interference on CP potential and performance were investigated.

Results demonstrated that the shift of CP potential depends on the CP level and AC current density.

iii

No matter if the DC potential of the steel is shifted negatively or positively upon application of

AC, the steel suffers from increased corrosion.

iv

Acknowledgements

I would like to express my sincere gratitude to my supervisor, Dr. Frank Cheng for his constant

guidance, encouragement, and support throughout my whole Ph. D program. His deep love and

perception of science, his persistent endeavour for searching for the truth, and his consistent efforts

at achieving perfection have always inspired and helped me carry out this research project.

Thanks are also extended to the members in my group, Drs. Luyao Xu, Ruijing Jiang and Tao

Liu, Dong Han, Qiang Li, and those whose names cannot all be listed here, for their helps and

valuable discussions in this work.

The generous financial supports from Canada Research Chairs Program and Eyes High

Doctoral Scholarship of the University of Calgary are highly appreciated, and make this work

possible.

v

Dedication

For

my parents and other family relatives,

and for others who have taught, guided and supported me in the past 28 years.

vi

Table of Contents

Abstract ............................................................................................................................... ii Acknowledgements ............................................................................................................ iv

Dedication ............................................................................................................................v Table of Contents ............................................................................................................... vi List of Tables .......................................................................................................................x List of Figures and Illustrations ......................................................................................... xi List of Symbols, Abbreviations and Nomenclature ........................................................ xvii

CHAPTER ONE: INTRODUCTION ..................................................................................1 1.1 Research background .................................................................................................1 1.2 Research objectives ....................................................................................................3

1.3 Content of thesis ........................................................................................................4

CHAPTER TWO: LITERATURE REVIEW ......................................................................6 2.1 Recognition of AC corrosion of pipelines .................................................................6

2.2 Principles of AC interference ....................................................................................7 2.2.1 AC interference sources ....................................................................................8

2.2.2 Capacitive coupling ...........................................................................................9 2.2.3 Resistive coupling ...........................................................................................10 2.2.4 Inductive coupling ...........................................................................................12

2.3 Fundamentals of AC corrosion ................................................................................14 2.3.1 Morphological characteristics .........................................................................14

2.3.2 AC corrosion mechanisms ...............................................................................17 2.3.3 Parametric effects on AC corrosion ................................................................24

2.4 CP interference by AC .............................................................................................30 2.4.1 Principle of impressed current CP ...................................................................30

2.4.2 Ineffectiveness of CP by AC ...........................................................................32

CHAPTER THREE: AC CORROSION AT COATING DEFECT ON PIPELINES .......37 3.1 Introduction ..............................................................................................................37

3.2 Experimental ............................................................................................................38 3.2.1 Specimen and solution .....................................................................................38 3.2.2 AC corrosion testing and DC potential derivation ..........................................39 3.2.3 Surface characterization ..................................................................................40

3.3 Results ......................................................................................................................40 3.3.1 AC corrosion testing on coated steel electrodes containing a defect with varied

sizes ..................................................................................................................40 3.3.2 AC corrosion testing on coated steel electrodes containing defects of varied

shapes ...............................................................................................................48 3.4 Discussion ................................................................................................................53

3.4.1 Effect of defect size on AC corrosion of coated steel .....................................53

3.4.2 Effect of defect shape on AC corrosion of coated steel ..................................60 3.5 Summary ..................................................................................................................61

vii

CHAPTER FOUR: AC INDUCED PITTING CORROSION ON PIPELINES IN BOTH

HIGH PH AND NEUTRAL PH CARBONATE/BICARBONATE SOLUTIONS ..........63 4.1 Introduction ..............................................................................................................63 4.2 Experimental ............................................................................................................64

4.2.1 Electrode and solution .....................................................................................64 4.2.2 AC corrosion and electrochemical measurements ..........................................64 4.2.3 Surface characterization ..................................................................................65

4.3 Results ......................................................................................................................66 4.3.1 AC corrosion testing in high pH solution ........................................................66

4.3.2 AC corrosion testing in neutral pH solution ....................................................70 4.4 Discussion ................................................................................................................73

4.4.1 AC corrosion of pipeline steel in high pH solutions .......................................73 4.4.2 AC corrosion of steel in neutral pH solution ...................................................76

4.5 Summary ..................................................................................................................78

CHAPTER FIVE: DEGRADATION OF EPOXY COATINGS ON PIPELINES IN THE

PRESENCE OF AC INTERFERENCE ............................................................................79 5.1 Introduction ..............................................................................................................79

5.2 Experimental ............................................................................................................80 5.2.1 Coatings, steel and solution .............................................................................80 5.2.2 Measurements of coating properties ................................................................81

5.2.3 Structural and morphological characterization ................................................83 5.3 Results ......................................................................................................................84

5.3.1 Coating disbonding tests ..................................................................................84 5.3.2 Water upktake tests ..........................................................................................87 5.3.3 EIS measurements ...........................................................................................90

5.4 Discussion ................................................................................................................93

5.4.1 Effect of AC on coating performance .............................................................93 5.4.2 Effect of AC on coating structure ....................................................................93

5.5 Summary ..................................................................................................................99

CHAPTER SIX: PROBING POTENTIAL AND SOLUTION PH UNDER DISBONDED

COATING ON PIPELINES IN THE ABSENCE OF AC INTERFERENCE ................100

6.1 Introduction ............................................................................................................100 6.2 Experimental ..........................................................................................................101

6.2.1 Electrode and solution ...................................................................................101 6.2.2 Simulated crevice cell ....................................................................................101

6.3 Results ....................................................................................................................103

6.3.1 Distribution of local potential under disbonded coating ...............................103 6.3.2 Distribution of solution pH under disbonded coating ...................................108

6.4 Discussion ..............................................................................................................112 6.4.1 Effect of applied potential on CP shielding ...................................................112

6.4.2 Effect of disbondment thickness on CP shielding .........................................113 6.5 Summary ................................................................................................................114

CHAPTER SEVEN: CATHODIC PROTECTION SHIELDING UNDER COATING

DISBONDMENT ON PIPELINES IN THE ABSENCE OF AC INTERFERENCE .....115

viii

7.1 Introduction ............................................................................................................115 7.2 Experimental ..........................................................................................................116

7.2.1 Coatings, steel and solution ...........................................................................116 7.2.2 Measurements of CP permeability of coatings ..............................................117

7.2.3 Coating characterization ................................................................................118 7.3 Results ....................................................................................................................118

7.3.1 Morphological observation of steel electrodes ..............................................118 7.3.2 Electrochemical measurements .....................................................................123 7.3.3 In-situ pH monitoring under disbonded coating ............................................127

7.3.4 FT-IR characterization of the coatings ..........................................................129 7.4 Discussion ..............................................................................................................133

7.4.1 Shielding effect of HDPE coating on CP ......................................................133 7.4.2 Permeability of FBE coating on CP ..............................................................134

7.5 Summary ................................................................................................................137

CHAPTER EIGHT: CATHODIC PROTECTION SHIELDING UNDER COATING

DISBONDMENT ON PIPELINES IN THE PRESENCE OF AC INTERFERENCE ...139 8.1 Introduction ............................................................................................................139

8.2 Experimental ..........................................................................................................140 8.2.1 Coatings, steel and solution ...........................................................................140 8.2.2 CP permeation measurements in the presence of AC interference ...............140

8.3 Results ....................................................................................................................141 8.3.1 Potential distribution under the coating disbondment ...................................141

8.3.2 Solution pH distribution under the coating disbondment ..............................145 8.4 Discussion ..............................................................................................................149

8.4.1 Effect of AC current density on coating disbondment ..................................149

8.4.2 Effect of AC current density on CP permeation into disbonding crevice .....151

8.5 Summary ................................................................................................................154

CHAPTER NINE: EFFECTS OF AC INTERFERENCE ON CATHODIC PROTECTION

POTENTIAL AND ITS EFFECTIVENESS FOR CORROSION PROTECTION ON

PIPELINES ......................................................................................................................155 9.1 Introduction ............................................................................................................155

9.2 Experimental ..........................................................................................................156 9.2.1 Electrode and solution ...................................................................................156

9.2.2 DC potential measurements ...........................................................................157 9.2.3 Weight-loss tests ............................................................................................158

9.3 Results ....................................................................................................................159

9.3.1 Measurements of DC potential of the steel ...................................................159 9.3.2 Measurements of CP current density .............................................................162 9.3.3 Weight-loss measurements ............................................................................165

9.4 Discussion ..............................................................................................................166

9.4.1 Effect of AC on shift of CP potential ............................................................166 9.4.2 Effect of AC on CP performance ..................................................................170

9.5 Summary ................................................................................................................171

CHAPTER TEN: CONCLUSIONS AND RECOMMENDATIONS .............................173

ix

10.1 Conclusions ..........................................................................................................173 10.2 Recommendations ................................................................................................175

REFERENCES ................................................................................................................177

x

List of Tables

Table 2.1. New CP criteria with consideration of AC current interference for protection of

buried pipelines [87]. ............................................................................................................ 35

Table 5.1. Chemical composition of extracted soil solution (unit: mg/L) .................................... 81

Table 5.2. Characteristic bands of epoxy coating in the FT-IR spectrum .................................... 96

Table 7.1. Characteristic functional groups of HDPE identified in the FT-IR spectrum in Fig.

7.8. ....................................................................................................................................... 130

Table 7.2. Characteristic functional groups of FBE identified in the FT-IR spectrum in Fig.

7.9. ....................................................................................................................................... 132

xi

List of Figures and Illustrations

Fig. 2.1. Corrosion rate of steels in soils versus AC current density [25]. .................................... 7

Fig. 2.2. Examples of (a) HVAC transmission lines and (b) AC traction system used in

railway. .................................................................................................................................... 8

Fig. 2.3. Schematic diagram of the capacitive coupling. ............................................................. 10

Fig. 2.4. Schematic diagram of the resistive coupling. ................................................................ 11

Fig. 2.5. Schematic diagram of the inductive coupling. .............................................................. 13

Fig. 2.6. Schematic illustration of a tubercle of ‘stone hard soil’ observed to grow from the

coating defect in connection with AC corrosion [35]. .......................................................... 15

Fig. 2.7. Leak site on underground natural gas transmission pipeline before (left) and after

(right) cleaning of deposit, where the arrow indicates the leak [39]..................................... 17

Fig. 2.8. Simplified AC corrosion process [41]. .......................................................................... 18

Fig. 2.9. A schematic illustration of the electrical equivalent circuit [35]................................... 19

Fig. 2.10. Schematic diagram of the alkalization mechanism [42]. ............................................. 20

Fig. 2.11. Pourbaix diagram: the hatched area indicates the critical AC corrosion zone [42]. ... 21

Fig. 2.12. Schematic diagram of double-charge layer of the steel/solution interface and the

AC induced corrosion model [46]. ........................................................................................ 23

Fig. 2.13. Maximum calculated induced voltage at various transmission line crossing angles

[55]. ....................................................................................................................................... 25

Fig. 2.14. (a) CP by galvanic anode and (b) CP by impressed currents [27]. .............................. 30

Fig. 2.15. Schematic diagram of the 'internal feedback' and 'external feedback' CP system

[17]. ....................................................................................................................................... 32

Fig. 2.16. New CP criterion based on AC and DC current densities [88]. .................................. 33

Fig. 2.17. New CP criteria with consideration of AC interference for pipeline steel in

carbonate/bicarbonate solution [87]. ..................................................................................... 35

Fig. 3.1. Schematic diagram of the experimental setup for AC corrosion in the simulated soil

solution. ................................................................................................................................. 39

Fig. 3.2. Time dependence of DC potential of the steel electrode containing a 20 mm

diameter defect at various AC currents in the test solution. ................................................. 41

xii

Fig. 3.3. Polarization curves measured on the coated steel electrode containing a 20 mm

defect under various AC currents in the solution. ................................................................. 42

Fig. 3.4. Optical views of the coated steel electrode containing a 20 mm defect after 12 h of

testing in the test solution at various AC currents (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20

mA. ........................................................................................................................................ 43


diameter defect at various AC currents in the test solution. ................................................. 44



Fig. 3.7. Optical views of coated steel containing a 10 mm defect in the diluted bicarbonate

solution at various AC currents after 12 h of test (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20

mA. ........................................................................................................................................ 45

Fig. 3.8. Time dependence of DC potential of the steel electrode containing a 5 mm diameter

defect at various AC currents in the test solution. ................................................................ 46




solution at various AC currents after 12 h of test (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20

mA. ........................................................................................................................................ 48

Fig. 3.11. Time dependence of DC potential of the steel electrode containing a circular,

square, and triangular defect, each with an area of 78.5 mm2 under 500 A/m2 AC current

density in the test solution. .................................................................................................... 49

Fig. 3.12. Polarization curves measured on the steel electrode containing a circular, square,

and triangular defect under 500 A/m2 AC current density in the test solution. .................... 50





and triangular defect, each with an area of 38.5 cm2 under 500 A/m2 AC current density

in the test solution. ................................................................................................................ 51




xiii


and triangular defect, each with an area of 19.6 cm2 under 500 A/m2 AC current density

in the test solution. ................................................................................................................ 53

Fig. 3.17. DC potential and polarization curves measured on the steel electrode where no AC

is applied, containing defects with different sizes in the test solution. ................................. 55

Fig. 3.18. DC potential of the steel electrode containing differently sized defects at AC

currents of (a) 1 mA; (b) 10 mA; and (c) 20 mA. ................................................................. 57

Fig. 3.19. Polarization curves measured on the steel electrode containing differently sized

defects at AC currents of (a) 1 mA; (b) 10 mA; and (c) 20 mA. .......................................... 59

Fig. 3.20. Morphology of steel at defect with 19.6 mm2 area under 500 A/m2 AC current


Fig. 4.1. Time dependence of DC (direct current) potential of X65 steel electrode at various

AC current densities in the high pH solution: (a) corrosion potential before 1200 s; (b)

AC-ON potential from 1200 s to 2400 s; (c) AC-OFF potential after 2400 s. ..................... 66

Fig. 4.2. Potentiodynamic polarization curves of the steel measured at various AC current

densities in high pH solution (a) 0 A/m2; (b) 100 - 500 A/m2. ............................................. 68

Fig. 4.3. Optical views of the steel electrode after different testing times in the high pH

solution under various AC current densities. From left to right: 5 min, 2 h and 12 h. The

magnification bar is 500 μm for all photos. .......................................................................... 69


AC current densities in the neutral pH solution: (a) corrosion potential before 1200 s; (b)

AC-ON potential from 1200 s to 2400 s; (c) AC-OFF potential after 2400 s. ..................... 70


densities in neutral pH solution. ............................................................................................ 71

Fig. 4.6. Optical views of the steel electrode after different testing times in the neutral pH


magnification bar is 500 μm for all photos. .......................................................................... 72

Fig. 4.7. Schematic diagram of the mechanistic model for AC corrosion of steel in high pH

solution, where the red box refers to the film. ...................................................................... 75

Fig. 4.8. Schematic diagram of the mechanistic model for AC corrosion of steel in neutral

pH solution, where the red box refers to the corrosion product layer. .................................. 77

Fig. 5.1. Schematic diagram of the experimental setup for CD test. ........................................... 82

Fig. 5.2. Experimental set-up for the test of permeability of coatings to water. .......................... 83

xiv

Fig. 5.3. Surface morphology of FBE coated steel specimen after cathodic disbonding test

under various AC current densities (a) 0, (b) 100 A/m2, (c) 200 A/m2, (d) 300 A/m2, (e)

400 A/m2, (f) 500 A/m2. ........................................................................................................ 84

Fig. 5.4. Coating disbonding radius as a function of AC current density measured in FBE

coated steel specimens. ......................................................................................................... 85

Fig. 5.5. Surface morphology of liquid epoxy coated specimens after CD test under various

AC current densities (a) 0, (b) 100 A/m2, (c) 200 A/m2, (d) 300 A/m2, (e) 400 A/m2, (f)

500 A/m2. .............................................................................................................................. 86

Fig. 5.6. Coating disbonding radius as a function of AC current density measured in liquid

epoxy coated steel specimens. .............................................................................................. 87

Fig. 5.7. Time dependence of the weight of the coating-sealed test container under various

AC voltages: (a) FBE coating; (b) liquid epoxy coating. ..................................................... 88

Fig. 5.8. Water transmission rate of the coatings under various AC voltages. ............................ 89

Fig. 5.9. Nyquist diagrams measured on coated steel specimens after 5 days of immersion in

the solution under various AC voltages: (a) FBE (b) liquid epoxy coating. ......................... 91

Fig. 5.10. Nyquist diagrams measured on coated steel specimens after 30 days of immersion

in the solution under various AC voltages: (a) FBE, (b) liquid epoxy coating..................... 92

Fig. 5.11. SEM images of the cross-sectional view of FBE coating after 30 days of

immersion: (a) without AC; (b) AC voltage of 50 V. ........................................................... 94

Fig. 5.12. SEM images of the cross-sectional view of liquid epoxy coating after 30 days of

immersion: (a) without AC; (b) AC voltage of 50 V. ........................................................... 94

Fig. 5.13. FT-IR spectrum of FBE coating after 30 days of immersion in the test solution

under various AC voltages. ................................................................................................... 95

Fig. 5.14. FT-IR spectrum of the liquid epoxy coating after 30 days of immersion in the test

solution under various AC voltages. ..................................................................................... 96

Fig. 5.15. Schematic diagram of the conceptual model for illustrating the coating

degradation in the presence of AC interference. ................................................................... 98

Fig. 6.1. Schematic diagram of the experimental setup simulating a disbonding crevice under

coating and the potential/solution pH measurements. ........................................................ 102

Fig. 6.2. Time dependence of the distributions of local potential under disbonded coating

(disbonding thickness of 120 μm) at varied disbonding depths from the open holiday

where the steel was either at (a) corrosion potential at (b) CP potential -0.875 V (SCE)

and (c) -0.975 V (SCE) respectively. .................................................................................. 105

xv

Fig. 6.3. Distributions of local potential under disbonded coating at varied disbonding depths

from the open holiday where the CP potential of -0.875 V (SCE) is applied under

various disbonding thickness (a) 120 μm, (b) 240 μm, (c) 360 μm. ................................... 107

Fig. 6.4. Time dependence of the distributions of local solution pH under disbonded coating

(disbonding thickness of 120 μm) at varied disbonding depths from the open holiday

where the steel was either at (a) corrosion potential at (b) CP potential -0.875 V (SCE)

and (c) -0.975 V (SCE) respectively. .................................................................................. 109

Fig. 6.5. Distributions of solution pH under disbonded coating at varied disbonding depths

from the open holiday where the CP potential of -0.875 V (SCE) is applied under

various disbonding thickness (a) 120 μm, (b) 240 μm, (c) 360 μm. ................................... 111

Fig. 7.1. Schematic diagram of the experimental setup to measure the permeability of

coatings to CP, where WE and RE refer to working electrode and reference electrode,

respectively. ........................................................................................................................ 118

Fig. 7.2. Optical morphology of the steel electrode in 0.01 M NaHCO3 solution trapped

under the disbonded HDPE coating after various times of testing at an applied CP

potential of -0.85 V (CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days. ............. 119


under the disbonded FBE coating after various times of testing at an applied CP

potential of -0.85 V(CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days. .............. 120


under the disbonded FBE coating after various times of testing at an applied CP

potential of -1.0 V (CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days. ............... 122

Fig. 7.5. The potential of steel electrode in 0.01 M NaHCO3 solution trapped under the

disbonded HDPE and FBE coatings under CP potentials of -0.85 V (CCS) and -1.0 V

(CCS), respectively: (a) HDPE, -0.85 V (CCS); (b) FBE, -0.85 V (CCS); (c) FBE, -1.0

V (CCS). ............................................................................................................................. 124

Fig. 7.6. Potentiostatic current measured on steel in 0.01 M NaHCO3 solution trapped under

the disbonded HDPE and FBE coatings under CP potentials of -0.85 V (CCS) and -1.0

V (CCS), respectively: (a) HDPE, -0.85 V (CCS); (b) FBE, -0.85 V (CCS); (c) FBE, -

1.0 V (CCS). ....................................................................................................................... 127

Fig. 7.7. Time dependence of pH of the solution under HDPE and FBE coatings at -0.85

V(CCS) and -1.00 V(CCS) CP potentials, respectively: (a) HDPE, -0.85 V(CCS); (b)

FBE, -0.85 V(CCS); (c) FBE, -1.0 V(CCS). ...................................................................... 129

Fig. 7.8. FT-IR spectrum of the HDPE coating before and after 30 days of testing under CP

of -0.85 V (CCS). ................................................................................................................ 130

Fig. 7.9. FT-IR spectrum of the FBE coating before and after 30 days of testing under CP of

-0.85 V (CCS). .................................................................................................................... 131

xvi

Fig. 7.10. Schematic diagram of the model for illustrating the compatibility of FBE coating

with CP. ............................................................................................................................... 135

Fig. 9.1. Schematic diagram of the experimental setup to measure the DC potential of the

steel electrode under various AC current densities and CP potentials. ............................... 158

Fig. 9.2. DC potentials of the steel electrode at various AC current densities under CP

potential of -0.850 V (CCS) in NS4 solution ...................................................................... 160


potential of -0.925 V (CCS) in NS4 solution ...................................................................... 161


potential of -1.0 V (CCS) in NS4 solution .......................................................................... 162

Fig. 9.5. CP current densities of the steel electrode at various AC current densities under CP

potential of -0.850 V (CCS) in NS4 solution. ..................................................................... 163


potential of -0.925 V (CCS) in NS4 solution. ..................................................................... 164


potential of -1.0 V (CCS) in NS4 solution. ......................................................................... 165

Fig. 9.8. Corrosion rate of the steel as a function of AC current density and CP potential

after 10 days of weight-loss testing in NS4 solution. ......................................................... 166

Fig. 9.9. Schematic diagram of the mechanistic model to illustrate the effect of AC on shift

of the DC potential and at CP potentials (a) -0.850 V (CCS); (b) -0.925 V (CCS); (c) -

1.0 V (CCS). ....................................................................................................................... 169

xvii

List of Symbols, Abbreviations and Nomenclature

Symbol Definition

A Exposed Area of Coating

AC Alternating Current

ASTM American Society for Testing and Materials

AV Alternating Voltage

C Capacitor

CD Cathodic Disbonding

CE Counter Electrode

CEPA Canadian Energy Pipeline Association

CP Cathodic Protection

CCS Cu/CuSO4 Electrode

DAQ Data Acquisition

DC Direct Current

DIN Deutsches Insitut Normung

EIS Electrochemical Impedance Spectroscopy

E01 Equilibrium Potential of Anodic Reaction

E02 Equilibrium Potential of Cathodic Reaction

FBE Fusion Bonded Epoxy

FT-IR Fourier Transform-Infrared Spectroscopy

GDP Gross Domestic Product

HDPE High Density Polyethylene

HVAC High Voltage Alternating Current

iAC Alternating Current Density

NACE National Association of Corrosion Engineers

Rs Spread Resistance

RE Reference Electrode

RMS Root Mean Square

SCE Saturated Calomel Electrode

SEM Scanning Electron Microscopy

VB1 Charge Transfer Resistance for Anodic Reaction

VB2 Charge Transfer Resistance for Cathodic Reaction

W Warburg Impedance

WE Working Electrode

WTR Water Transmission Rate

d Diameter of a Circular Coating Defect

ΔG Weight-loss

1

Chapter One: Introduction

1.1 Research background

Pipelines are considered as lifelines of the global energy industry. Compared to other modes

of transportation like marine shipping, railroad, aircraft and truck transport, pipelines provide safe,

reliable, efficient, and economical means for energy transportation [1]. In Canada, a network of

approximately 117,000 kilometres of underground energy transmission pipelines operates every

day transporting crude oil, diluted bitumen and nature gas from the production sites to refineries

and the markets, contributing to a strong national economy. According to Canadian Energy

Pipeline Association (CEPA) [2], transmission pipeline operations of all types added more than

$8.8 billion to Canada’s gross domestic product (GDP) in 2012.

Safety is the top priority of pipeline systems. However, there has been a wide variety of threats

to pipeline integrity [3], including material and construction defects, mechanical damage from

construction, maintenance or third-party excavation, incorrect cathodic protection operation,

corrosion and coating degradation, geotechnical forces, etc. In particular, with the increasing

collocation of pipelines and high-voltage electric power lines, the alternating current (AC)

interference has been recognized as a major safety risk to pipelines.

When a buried pipeline parallels or crosses a high-voltage overhead power line, the flow of

AC in the power line generates an alternating magnetic field in the air and soil surrounding, then

in turn generates an inductive AC on the pipeline. If the pipeline coating contains microscopic or

macroscopic defects, the steel at these areas will be subject to AC-induced corrosion. AC corrosion

has been recognized since the beginning of the 20th century [4-6]. However, for many years,

corrosion engineers did not consider corrosion due to AC on metallic structures to be an important

2

phenomenon. The prevailing opinion was that, although AC could cause corrosion of steel, the

corrosion rate was a small percentage of an equivalent amount of direct current (DC) and the

interference could be controlled by the application of cathodic protection (CP). AC corrosion has

received extensive interest since the first corrosion damage induced by AC on cathodically

protected pipelines found in Germany in 1986. In the past 30 years, many pipeline failures in North

America and Europe have been ascribed to AC interference even when CP was applied [7, 8].

Typical examples include the AC induced corrosion on a section of a cathodically protected 412

km, 10 inch ethylene pipeline running through Scotland and England [9] and an AC induced leak

on a cathodically protected natural gas pipeline located on south of Oswego in America [10].

Generally, the buried pipelines are protected from corrosion attack by coatings and CP.

However, pipeline coatings usually contain various defects such as pinholes, holidays or flaws,

which are frequently generated during coating manufacturing, transportation and pipeline

construction. Ground water, corrosive gases and chemical species may penetrate through the

defects and reach the pipe steel surface, resulting in coating disbonding and pipeline corrosion at

the defect and under the disbondment. To date, there has been few work to study the effect of AC

on coating disbondment and the coating properties.

Furthermore, the AC interference affects the performance of CP system and deviates the

potential of the protected pipeline from the applied value, reducing the CP effectiveness for

corrosion protection [11]. For example, it was recognized that the National Association of

Corrosion Engineers (NACE) recommended CP criterion is not able to provide a full protection

for pipelines in the presence of AC interference [12]. Moreover, field tests have demonstrated that

AC-induced corrosion can occur on carbon steels even under higher CP levels [13-15]. It was also

found that an excessive CP could increase the AC corrosion rate [16]. Obviously, there has been

3

no consensus on the role of AC in CP performance for corrosion protection. At the same time, it

was confirmed [17] that the shift of CP potential by AC depends on both the AC current density

and the CP level. At CP potential of -0.85 V (saturated calomel electrode, SCE), the AC application

shifts the direct current (DC) potential of the pipe steel negatively. When the CP potential is -1.0

V (SCE), the DC potential of the steel is positively shifted with AC. The exact mechanism for the

AC driven shift of the steel potential has remained unclear.

In addition to the adverse effects on coating performance and CP effectiveness, the AC is able

to accelerate pipeline corrosion and result in localized pitting corrosion at high AC current

densities. However, the critical AC current density to initiate pitting and the mechanism of AC

induced pitting corrosion on steels have so far remained unknown. Furthermore, the accelerating

AC corrosion and pitting corrosion of buried pipelines are highly dependent on the presence of

coating defects and their geometry such as the size and shape of the defect [18], but limited

understanding has been achieved in this area.

1.2 Research objectives

The overall objective of this research is, through experimental testing and mechanistic

modeling, to advance the fundamental understanding of pipeline external corrosion and coating

degradation in the presence of AC interference. Progress will be made in the following areas.

1) To investigate the effect of the geometry of coating defects on AC corrosion and understand

mechanistically the role of coating defects in pipeline corrosion under AC interference.

2) To investigate the mechanistic aspects of AC induced pitting corrosion on pipelines.

3) To study the effect of AC on properties and performance of epoxy coatings.

4

4) To investigate the essential effect of AC on CP shielding under disbonded coatings by in-

situ probing the distributions of local potential and solution pH.

5) To investigate the effect of AC on shift of CP potential and derive the ‘true’ potential

applied on pipelines in the presence of AC interference.

1.3 Content of thesis

The thesis contains ten chapters, with Chapter One giving an overall introduction of the

research background and objectives.

Chapter Two reviews comprehensively the principle of AC interference on pipeline integrity,

including the fundamentals of AC induced corrosion and the recognition of AC interferences on

the applied CP as well as coating performances.

Chapter Three is devoted to studying the effect of coating defects, including their size and

geometry, on pipeline corrosion under AC interference.

Chapter Four investigates the AC induced pitting corrosion of X65 pipeline steel in both high

pH and neutral pH solutions that are encountered under disbonded coatings on pipelines. A

mechanistic model is developed to illustrate the pitting initiation and growth under AC

interference.

Chapter Five studies the effect of AC on the molecular structure and properties of epoxy

coatings in a simulated soil solution by measuring the adhesion, water uptake and corrosion

resistance, as well as the morphological and structural characterization of the coatings upon AC

application.

5

Chapter Six monitors the local potential and solution pH under disbonded coating on X65

pipeline steel and determines the shielding effect of coating disbondment on CP in the absence of

AC interference.

Chapter Seven investigates the CP permeation through coating membranes by measurements

of DC potential, potentiostatic current density, and coating molecule characterization.

Chapter Eight studies the role of AC in CP permeation into the coating disbondment through

electrochemical measurements and pH monitoring. Parametric effects such as AC current density,

CP potential and disbonding thickness, on CP permeation are determined.

Chapter Nine analyzes the effect of AC interference on shift of CP potential. A model is

developed to illustrate the essence and dependence of the steel potential on applied CP potential

in the presence of AC interference.

Chapter Ten summarizes the main conclusions of this research, along with recommendations

for the further work.

6

Chapter Two: Literature Review

2.1 Recognition of AC corrosion of pipelines

Studies of pipeline corrosion induced by AC can be dated back to the late 19th century. For

many years, corrosion engineers did not consider enhanced corrosion attributed to AC on metallic

structures important. Up until the mid-1980s, the prevailing opinion was that, although AC could

cause corrosion of steels, the corrosion rate was a small percentage of an equivalent amount of

direct current (DC) and could be controlled by the application of CP in accordance with the

protection criterion. AC corrosion was not paid proper attention for two reasons: (1) the interaction

of AC and DC currents affecting the corrosion is very complicated; and (2) the instruments

normally used to measure the electric parameters in DC cannot correctly detect the presence of AC

with frequencies between 50 and 100 Hz [19, 20]. Recently, there has been an increasing concern

for AC corrosion since AC interference has been demonstrated to affect catholically protected

pipelines and induce safety issues. The late 'wake-up' of the industry is due to factors including the

growing number of high-voltage transmission lines and high AC current densities generating at

coating defects along the pipeline [21]. Nowadays, AC corrosion has been recognized as a serious

threat to the integrity of underground structures, including pipelines.

There is a scarcity of data on the corrosion rate of pipeline steels in the presence of AC

interference. The general understanding is that a higher AC leads to higher risk of AC corrosion

of the steels [22]. An AC corrosion case of a gas pipeline located under a high voltage AC utility

corridor was reported near a power plant of Rainier, Oregon in 2001 [23]. Four holes were

observed in the pipeline after only five months in use under the AC voltage of 90 V, indicating a

corrosion rate of over 400 mpy. Estimations showed [24] that, in soils with pH between 6 and 7,

7

the corrosion rate of steels is below 0.07 mm/y at AC current density of 160 A/m2 and between

0.25 and 0.50 mm/y at 780 A/m2. The result of corrosion rate data from literature of cathodically

protected steels in the presence of AC interference is shown in Fig. 2.1 [25]. It is seen that the

corrosion rate could reach up to 2.5 mm/y under the AC current density of about 300 A/m2. A

German field-based coupon study [26] observed the pitting corrosion rate of 5.3 mm/y associated

with AC densities between 20 and 200 A/m2. The long term progression of the corrosion rate over

time has been still unknown.

Fig. 2.1. Corrosion rate of steels in soils versus AC current density [25].

2.2 Principles of AC interference

Generally, occurrence of electric interferences requires the existence of a source of disturbance,

a coupling mechanism and a receptor. In the case of AC interference, the source of disturbance is

8

the power line or the AC traction system, the receptor is the metallic structure such as a pipeline,

and the coupling between the power line and the pipeline occurs by a capacitive, resistive or an

inductive mechanism [27]. The increased number of AC corrosion related failures is associated

with the growing parallelism between buried pipelines and high voltage transmission lines which

share the same right of way.

2.2.1 AC interference sources

The main AC interference sources are high-voltage AC (HVAC) transmission lines and AC

traction systems (usually fed by a parallel high voltage line at 50 Hz or 16.7 Hz), as shown in Fig.

2.2. The HVAC lines are used to transmit electric power over relatively long distances or used as

electric power transmission from one central station to another for load sharing. The HVAC lines

are made of high voltage (between 138 and 765 kV) overhead or underground conducting lines of

either copper or aluminium. Transmission lines, when interconnected with each other, become the

high voltage transmission network.

Fig. 2.2. Examples of (a) HVAC transmission lines and (b) AC traction system used in

railway.

(a) (b)

9

Transmission lines mostly use a three-phase AC system, while single phase AC is sometimes

used in railway electrification systems. Aluminum alloy reinforced with steel strands is usually

used as the conductor material due to the advantages in weight and cost compared with copper. In

the transmission network, transmission substations are used to decrease the voltage of high

incoming electricity by means of voltage transformers, allowing the connection of HVAC lines

and the lower voltage distribution network. Finally, at the point of use, the energy is transformed

to lower voltage (varying by country and customer requirements).

There is a wide variety of electric traction systems around the world, which have been built

according to the type of railway, location and the available technology at the time of installation.

Electric railway networks can use either alternating or direct current. The AC systems always use

overhead wires, while DC systems can use either an overhead wire or a third rail. The AC power

transmission system along the line is used mainly for long distance, while DC, on the other hand,

is the preferred option for shorter lines, urban systems and tramways. Compared with DC traction

system, it is much more convenient to increase AC voltage than DC voltage; thus, it is easier to

supply more electricity with an AC power system. The 25 kV AC railway electrification system is

commonly used in railways worldwide, especially on high-speed lines. The choice of 25 kV is

related to the efficiency of power transmission as a function of voltage and cost [28].

2.2.2 Capacitive coupling

The capacitive coupling is due to the influence of two or more circuits upon one another,

through a dielectric medium like air, by means of the electric field acting between them [29]. Fig.

2.3 shows the schematic diagram of the capacitive coupling. The electric field associated with

power conductors causes a current flow between a nearby aboveground metallic structure and the

http://en.wikipedia.org/wiki/Three-phase_electric_power




http://en.wikipedia.org/wiki/Single-phase_electric_power

http://en.wikipedia.org/wiki/Single-phase_electric_power

http://en.wikipedia.org/wiki/Railway_electrification_system

http://en.wikipedia.org/wiki/Railway_electrification_system

http://en.wikipedia.org/wiki/High-voltage_direct_current

http://en.wikipedia.org/wiki/High-speed_rail



10

earth. And this capacitive coupling usually induce an AC interference on aboveground sections of

the pipeline during installation. The magnitude of the total stray current is a function of the size of

the structure, the distance to power conductors, the voltage of power conductors, and their

geometrical arrangement.

Fig. 2.3. Schematic diagram of the capacitive coupling.

2.2.3 Resistive coupling

The resistive coupling is due to the influence of two or more circuits on one another by means

of conductive paths (metallic, semi-conductive, or electrolytic) between the circuits [29]. It is the

case of the grounded structure of an AC power system that shares the earth with other buried

structures. Coupling effects could transfer AC to a buried pipeline in the form of alternating current

or voltage.

11

Whenever a grounded power system has unbalanced condition, current may flow to the earth.

Indeed, the resistive coupling is primarily a concern during a short-circuit condition on an AC

power system—for instance, when a large part of the current in a power conductor flows to the

earth by means of foundations and grounding system of a tower or a substation. The current flow

raises the electric potential of the earth near the structure, often to thousands of volts with respect

to remote earth, and can result in a considerable AC voltage across the coating of a buried pipeline.

This can damage the coating and the structure itself. The potential difference between the earth

and the buried pipeline can represent an electric shock hazard. Under some circumstances, the

electric potential of the structure may be raised enough to transfer hazardous voltages over

considerable distances. The schematic diagram of the resistive coupling is shown in Fig. 2.4.

Fig. 2.4. Schematic diagram of the resistive coupling.

12

Lightning strikes to the power system can also initiate fault current conditions [29]. Lightning

strikes to a structure or to earth in the vicinity of a structure can produce electrical effects similar

to those caused by AC fault currents. If a lightning strike occurs between the tower structure and

an overhead cloud, the potential of the tower could be raised to an extremely high voltage, with a

current flow from the tower structure to the earth. The current flows to the earth through the

grounding system and spreads uniformly through the earth (assuming a homogenous soil

resistivity) in all radial directions. Therefore, the ground around the pipeline will be at a relatively

high potential with respect to the pipeline potential and this could result in the coating damage due

to heating effect. Fault currents can also enhance the coating disbondment and induce the further

corrosion beneath the disbondment. Resistive coupling effects are mainly dependent on the

following important factors [29]:

1) The total short-circuit current;

2) The electrical resistivity of the soil;

3) Separation distance between power systems and the affected buried pipelines;

4) The conditions of the pipeline coating.

2.2.4 Inductive coupling

The main coupling between the power line and the pipeline is occurred by an inductive

mechanism, which is due to the influence of two or more circuits upon one another by means of

the magnetic flux that links them [29]. AC flow in a power conductor produces an alternating

magnetic field around it. When a pipeline is close enough and parallel to the electrical transmission

line, the magnetic field will cross the pipeline with the induction of AC voltage on it. The

schematic diagram of the inductive coupling is shown in Fig. 2.5. The current flow in the conductor

13

creates an electromagnetic field that lies at right angles to the current that produces it. Then the

magnetic field expands away from the conductor and collapses towards the buried pipeline [30].

The flowing out of the AC current could induce corrosion under the coating defect or disbondment.

The magnitude of the induced potential depends on the following factors [29]:

1) The overall separation distance between the buried pipeline and the power line;

2) The length of exposure and the power line current magnitude;

3) The type of conductor used on the power line;

4) The coating resistance of the structure;

5) The grounding present on the structure;

6) The soil resistivity as a function of depth.

Fig. 2.5. Schematic diagram of the inductive coupling.

14

Numerical modeling can be utilized to examine the collocated pipeline’s susceptibility to

HVAC interference, helping identify locations of possible AC current discharge, and design

appropriate mitigation systems to reduce the effects of induced AC voltage, fault currents, and AC

corrosion to meet accepted industry standards [31]. Recently, a simulation software suite has been

presented to model both resistive and inductive interference between pipelines and power

transmission networks [32]. Moreover, a transient simulation of the currents along the towers and

ground wires is performed when the lightning strikes the tower of the power transmission line, and

methods are proposed to evaluate the safety distance between the pipeline and tower grounding

structure [33].

2.3 Fundamentals of AC corrosion

As stated above, a buried pipeline which collocates with HVAC transmission lines can be

interfered by the AC source. In the presence of AC interference, a so-called AC corrosion can

occur on the pipeline [34].

2.3.1 Morphological characteristics

A distinct tubercle of 'stone-hard soil', comprising a mixture of corrosion products and soil,

was observed at coating defects when AC corrosion incidents occur [35]. Fig. 2.6 shows the

schematic illustration of the characteristic tubercle. The specific resistivity of such a tubercle is

lower than that of the surrounding soil. The area of the tubercle is considerably greater than the

size of the original coating defect.

15

Fig. 2.6. Schematic illustration of a tubercle of ‘stone hard soil’ observed to grow from the

coating defect in connection with AC corrosion [35].

Recently, Ragault [36] has summarized 31 AC corrosion cases on polyethylene coated gas

pipelines. It was found that the corrosion products mainly consisted of magnetite mixed with soil.

Meanwhile, Williams [37] indicated that the corrosion product of steels under AC interference was

mostly magnetite. Wakelin et al. [25] described the Canadian AC case histories and summarized

the characteristic morphology of AC corrosion on pipelines, i.e., a rounded bottom pit and the hard

dome over the pit formed by corrosion products. Moreover, Bolzoni et al. [38] reported that the

AC led to the generation of thick but non-adhesive corrosion products.

An AC induced pipeline failure was investigated and the typical AC corrosion morphology

was reported [39]. The leak was inspected in a natural gas transmission pipeline (25 bar pressure,

150 mm inner diameter, 4.5 mm wall thickness, and 45 years in service) during a routine

16

inspection. It was found that the leak was covered by a large cap (approximately 200 mm diameter)

of hard, agglomerated soil, as shown in Fig. 2.7. The pipeline ran parallel to a railway (16.7 Hz

system) for approximately 10 km, and the leak occurred at the end of this section. The pipe was

coated with bitumen. Measurements after the failure indicated that AC voltage between 20 to 30

V was applied on the affected region. It was further found that the bituminous coating became

rather soft and sticky in the area surrounding the leak. After cleaning, localized attacks with deep

cavities in a generally passive surface were found, as shown in Fig. 2.7. The corrosion products

above the corroded area were analysed, and three types of materials were detected:

1) Material A: A black material based on bitumen with different amounts of iron oxide

(mainly magnetite) and a sodium compound. Inorganic components were finely dispersed

in the bituminous matrix;

2) Material B: A grey product based on the soil material, i.e., a silicate incorporating minor

amounts of alkali (Na, K) and earth-alkali (Ca, Mg) ions;

3) Material C: A white substance identified as a mixture of NaHCO3 and Na2CO3, without

any significant impurities.

In this case, AC interference was found to induce pitting corrosion of the pipeline steel and

cause the degradation of bitumen coating with deforming like a bubble and softening due to the

AC heating effect. Some specific features of AC induced corrosion of pipelines include:

The characteristic morphology of ‘stone hard soil’;

Pitting corrosion with localized attack on the pipeline steel;

Finely dispersed magnetite as the predominant corrosion product;

Softening coating; and

Increased pH near the corroded area.

17

Fig. 2.7. Leak site on underground natural gas transmission pipeline before (left) and after

(right) cleaning of deposit, where the arrow indicates the leak [39].

2.3.2 AC corrosion mechanisms

Various theories and models have been proposed to illustrate the mechanism of AC corrosion

of pipelines. However, none of them are able to fully explain the phenomenon and associated

phenomena.

Yunovich and Thompson [40] analyzed the reversibility of cathodic and anodic processes and

stated that the process occurring during the anodic half cycle of the AC signal might not be

completely reversed during the cathodic half cycle. As a result, the AC corrosion of metals is

attributed to the fact that the charges during the anodic cycle (metal dissolution) are greater than

those during the cathodic cycle (metal deposition). A simplified description of the AC corrosion

mechanism is schematically illustrated in Fig. 2.8 [41]. During the positive half wave, the metal is

oxidized, resulting in the formation of an oxide film. During the negative half wave (cathodic

behavior), the film is reduced to iron hydroxide. Then a new film grows in the following anodic

cycle. The amount of iron hydroxide increases due to the further reduction of the oxide film.

18

Hence, every AC cycle could result in the net oxidation of the metal and cause a significant metal

loss in the long term of AC interference. The reactions involved in the anodic and cathodic

processes are as follows:

3Fe + 4H2O → Fe3O4 + 8H+ + 8e (2.1)

Fe3O4 + 4H2O + 2e → 3Fe(OH)2 + 2OH- (2.2)

The model gives a brief but clear introduction of AC corrosion process, although it does not

consider the condition of CP application on the metal, and the influences of AC on electrochemical

behavior of the metal, both thermodynamically and kinetically.

Fig. 2.8. Simplified AC corrosion process [41].

19

Nielsen and Cohn [35] proposed an electrical equivalent circuit model to help understand the

AC corrosion process and mechanism. It is seen from Fig. 2.9 that a DC (represents the CP system)

and an AC source are applied individually between the pipeline and remote earth at a specific

location or coating defect. Soil resistance or spread resistance (Rs) affects the current flow through

the soil from remote earth to the coating defect. E01 and E02 are the equilibrium potentials of

anodic and cathodic reactions, respectively. The charge transfer resistances for each single reaction

are presented with diode elements, VB1 and VB2. Moreover, diffusion impedance (symbolized by

W, i.e., Warburg impedance) is attached to each reaction to restrict the rate by which the reactions

are rate-limited by diffusion processes of reactants. The interface between the steel and the

electrolyte acts as a capacitor, whose capacitance is the double layer capacitance, C. The key

element of the corrosion process is the VB1 element associated to metal dissolution. Corrosion

occurs if the electrical charge during the anodic half-cycle exceeds the cathodic charge.

Fig. 2.9. A schematic illustration of the electrical equivalent circuit [35].

20

Another widely accepted mechanism is the alkalization mechanism [42], where the combined

action of high pH and potential oscillation caused by AC interference induces corrosion attacks.

As shown in Fig. 2.10, hydroxide (OH-) produced by CP current accumulates in the near

surroundings of coating defects. This accumulation process needs an 'incubation period', which is

defined as the time to reach a critical pH value in the electrolyte in contact with the metal. Potential

fluctuations between the passivity, the immunity and high-pH corrosion regions in Pourbaix

diagram, as shown in Fig. 2.11, caused by the AC interference may induce corrosion as a result of

different time constants associated with iron dissolution (fast) and subsequent formation of passive

film (slow). At a very alkaline pH, the formation of dissolved HFeO2- may stabilize corrosion at a

very high rate.

Fig. 2.10. Schematic diagram of the alkalization mechanism [42].

21

Fig. 2.11. Pourbaix diagram: the hatched area indicates the critical AC corrosion zone

[42].

A modified two-step AC corrosion mechanism [43] was proposed based on the alkalization

theory. In step 1, AC causes breakdown of the passive film formed on steel surface due to electro-

mechanical stresses in the film. In step 2, corrosion occurs if the pH of the environment at the

metal-electrolyte interface is close to 14. However, it was not clear why the passive film

breakdown was caused by AC and there was no sufficient experimental results to demonstrate

whether the pH at the metal-electrolyte interface could reach up to 14.

Panossian et al. [44] suggested a similar mechanism that AC corrosion is a consequence of

oscillations of metal/electrolyte interface, which depends on the electrolyte pH. The AC corrosion

occurs due to the irreversibility of iron dissolution reaction and the impossibility of formation of a

22

passive film in alkaline electrolytes (pH < 14). Under excessively alkaline electrolytes (pH > 14),

the formation and reduction of a thin oxide film on metal surfaces could induce the AC corrosion.

In addition, Lalvani and Lin [45] proposed a revised model to introduce the effect of the double

layer capacity in the AC corrosion, but ignored the resistance of the electrolyte or the spread

resistance of the soil.

Xu et al. [46] developed an improved double-charge layer model to illustrate mechanistically

the AC corrosion of pipeline steels, as shown in Fig. 2.12. Charge-transfer reactions occur

primarily in the double-charge layer due to its extremely high strength of electric field, which

could be disturbed remarkably by AC. However, the majority of AC current flowing through the

double-charge layer acts as a non-Faradic charging-discharging current, which is usually much

higher than the Faradic current for reductive or oxidative reaction. The AC application could

generate a ‘vibrating’ effect due to the low frequency alternating electric field force exerting on

the reactants in the solution. The mobility rate of solvated ions, e.g., Fe2+, OH-, H+, etc. is sensitive

to the low frequency electric field force. Upon application of a low frequency AC (60 Hz in this

work), there is a sufficient time for ions to be accelerated by the electric field force, and

consequently, shortening the ‘free path’ for the ionic collision. This would increase the kinetic

energy of reactants and the opportunity of reactions among reactants. Corrosion is thus enhanced.

Furthermore, the effect of the evolution of hydrogen and oxygen by applied AC on the steel

corrosion is not negligible. The hydrogen atoms generated during corrosion reaction and

penetrated into steel is capable of enhancing corrosion activity of the steel. Moreover, the influence

of the oxygen evolution on corrosion of the steel is to supply more cathodic reactants, i.e., oxygen,

for the reductive reaction.

23

Fig. 2.12. Schematic diagram of double-charge layer of the steel/solution interface and the

AC induced corrosion model [46].

As the AC induced pitting corrosion is more often observed in the pipeline failure cases, studies

have tried to explain the development of pits in the presence of AC. Lalvani and Zhang [47]

conducted experiments in a 3.5 wt% NaCl solution at different applied sinusoidal AC signals, and

found an increase in pitting corrosion rates with time. Wen et al. [48] found that a higher AC

current density leads to an increasing pitting numbers as well as corrosion area on the surface of

pipeline steels. The relationship between the pit number and AC current density follows a power

function approximately. Fu et al. [49] discovered when the AC is as high as 500 A/m2, passivity

cannot maintained on the steel and pitting corrosion occurs extensively on the electrode surface in

a chloride-containing carbonate/bicarbonate solution.

24

An AC induced pitting corrosion mechanism [50] has been proposed based on the observation

of the auto-catalytic nature of pits. An excess concentration of positive charge due to metal

dissolution would accumulate inside the defect of the passive oxide layer. Negatively charged

chloride ions would migrate into the defect to maintain electroneutrality. Metal ions and chloride

ions will combine, resulting in large quantities of FeCl2. Through hydrolysis, a high concentration

of H+ ions will be present. The aggressive Cl− and H+ ions will develop an acidic environment, and

self-propagating pit will be formed in this defect. However, in this model, the initiation and

propagation of pits are dependent on the effect of chloride ions instead of the AC interference.

Cases of AC induced pitting corrosion of pipelines have been observed, where there is no chloride

ions present.

Obviously, there has still been no single mechanism that could give convincible explanation

of pipeline corrosion under AC interference, especially the AC induced pitting corrosion. It was

proposed that probably more than one theory is eligible for consideration of AC corrosion [51-53]

and a ‘mixed’ mechanism may exist for this phenomenon.

2.3.3 Parametric effects on AC corrosion

A great number of factors can influence AC corrosion of pipelines, as listed below.

Orientation of the pipeline to HVAC electric lines

The orientation of the pipeline relative to HVAC lines is critical in determining the magnitude

and distribution of induced AC voltage along the pipeline [54]. When the crossing angle between

the pipeline and the electric line increases, the magnitude of the induction decreases since the

section of the pipeline exposed to induction reduces. For a perpendicular crossing, with the

25

pipeline crossing at or near 90° to the AC power lines, the induction on the pipeline is minimized

as the effective parallel length is minimized. A numerical model [55] was developed to simulate a

6-inch pipeline coated with FBE in 10 Ω∙m soil running under a 345 kV AC transmission tower

geometry. Fig. 2.13 shows the results of the modeled maximum induced potential at various

crossing angles between 15 and 90 degree. Considering current load of up to 1000 A, it is seen that

a crossing angle of greater than 45 degree would induce about 2 V of AC potential, and that a

crossing of greater than 60 degree induces less potential such that the corresponding current

density is less than 20 A/m2 even in a relatively low soil resistivity at 1000 Ω∙cm.

Fig. 2.13. Maximum calculated induced voltage at various transmission line crossing

angles [55].

26

Industry experiences suggest that crossings of greater than 60 degree are often not considered

serious in terms of AC corrosion occurrence [56]. However, the simulation shows that the crossing

angle of 80 degree may even be considered significant in accelerating AC corrosion under a high

current load of 3000 or 5000 A. The modeling presents a noteworthy change from previous

accepted understanding [55].

AC frequency

AC corrosion could be induced under the frequency of 50 or 60 Hz. There is an agreement

about the effect of the AC frequency on AC corrosion, i.e., the corrosion rate of pipelines decreases

by the increasing frequency [57-59]. This is due to the fact that, with the increase of the AC

frequency, the interval between successive anodic and cathodic half-cycles becomes shorter.

Consequently, the metallic ions formed in the anodic cycle would be available for immediate re-

deposition in the cathodic cycle. In addition, at high frequencies, hydrogen atoms formed during

the cathodic cycles would not have sufficient time to coalesce and form hydrogen molecules.

During the next anodic half-cycle, a layer of hydrogen atoms cover the metal surface to prevent

the metal from dissolution reaction [60]. Moreover, it has been found that the AC frequency affects

pit morphology, pit density and passive current density. Generally, the passive current density of

steels in the electrolyte decreases with increasing AC frequency [61].

AC voltage

The technical specification CEN/TS 15280 [62] considers AC voltage an important parameter

to evaluate AC corrosion, as it can be directly measured on the structure. This specification

27

recommends that, in order to reduce AC corrosion, the AC voltage on pipelines should not exceed

at any time:

10 V where the local soil resistivity is greater than 25 Ω∙m, and

4 V where the local soil resistivity is lower than 25 Ω∙m.

The NACE standard [29] reports that a steady-state AC voltage of 15 V or more with respect

to local earth could constitute a shock hazard to personnel. However, the assessment of AC

corrosion threat based on the basis of AC voltage only may be misleading and not reliable. Many

other factors, such as coating defects and soil properties, should be taken into account to define

AC corrosion. Particularly, the parameter of AC current density should be used to characterize the

AC corrosion and gives more precise assessments [63].

AC current density

Both field experiences and lab testing have indicated that AC corrosion of pipelines occurs

above a critical AC current density [64]. Above the minimum AC current density, corrosion can

occur even in the presence of applied CP. The relevant AC current density criteria obtained in the

early investigations include [65]:

1) AC corrosion does not occur at AC current density less than 20 A/m2;

2) AC corrosion is unpredictable for AC current density between 20 to 100 A/m2;

3) AC corrosion occurs at current density greater than 100 A/m2.

Whereas there is a threshold AC current density, below which AC corrosion is not a concern,

the prevailing standards have a disputation on the magnitude of this critical value. The German

standard DIN 50925 [66] and the European standard CEN/TS 15280 [62] adopt the value of 30

A/m2, which could be detrimental to buried steel structures. Handbook of cathodic corrosion

28

protection [67] cites the value of 20 A/m2 as the threshold of AC current density to induce

corrosion. However, recent experimental study [68] showed that there likely was not a theoretical

'safe' AC current density. Goldanich et al. [69] found that the AC current density of 10 A/m2 may

be hazardous to pipelines, as it has increased the corrosion rate by two-folds in a simulated soil

solution compared to the AC-free conditions. In addition to enhance the general corrosion, AC is

able to result in localized corrosion under a high AC current density. Extensive corrosion pits were

observed on the surface of steel when the AC current density is increased to 800 A/m2 in a neutral

pH bicarbonate solution [17]. The critical AC current density inducing pitting corrosion of steels

has so far remained unknown.

Coating defects

The importance of a high quality and well-applied external pipeline coating is obvious for

corrosion protection. Coating systems are available today with exceptional dielectric and

mechanical characteristics that often result in very high coating resistances, especially in high soil

resistivity [70]. It is generally recognized that there are many defects in different sizes in the

coating due to manufacturing, careless construction, aging or damage during operation [18]. The

geometry of the coating defect, including the size and shape, has important influence on AC

corrosion. It has been found [71] that the smaller the defect, the greater the AC current density is.

The CP current is usually more difficult to reach a small size defect than a large defect nearby [72].

A number of field investigations indicated that the majority of corrosion occurs at holidays with

an area of 1 cm2 or smaller [73]. Jiang et al. [18] found that the presence of different sized defects

on the pipeline coating is important in AC corrosion occurrence.

29

Soil properties

The AC current density (iAC) at a coating defect depends on the induced AC voltage on the

pipeline and soil resistivity by [74]:

d

Vi AC

AC

8 (2.3)

where 𝜌 is soil resistivity and d is the diameter of a circular coating defect, which has a surface

area equal to that of a real holiday. This formula is applicable to cases when the defect size is

greater than the thickness of the coating [74]. It can be concluded that AC current density varies

linearly with AC voltage and depends on soil characteristics by its resistivity. The local soil

resistivity is controlled by the amount of soluble salts as well as water content, and is strongly

influenced by the electrochemical process occurring on the metal surface under CP. Depending on

the soil composition, the electrical resistance of the soil near the coating defect can either increase

or decrease with time due to chemical modifications when the pH is increased by the applied CP.

In particular, earth-alkaline ions (such as Ca2+ and Mg2+) form hydroxides with relatively low

solubility, thus significantly increasing the resistance at the coating defect. Otherwise, alkaline

cations (such as Na+, K+ or Li+) form highly soluble hygroscopic hydroxides, which can decrease

the electric resistance at the coating defect. Therefore, AC current density at a coating defect is

dependent on the ratio of alkali to earth-alkali ions, and the amount of hydroxyl ions, in addition

to the size of the coating holiday and specific soil resistivity.

30

2.4 CP interference by AC

2.4.1 Principle of impressed current CP

CP is an electrochemical method of corrosion prevention and protection which can be applied

to metals exposed to corrosive environments [75]. As shown in Fig. 2.14, this technique is based

on the circulation of direct current between an electrode (anode) placed in the environment and the

metallic structure (cathode). The cathodic current lowers the potential of the metal and reduces (or

halts) its corrosion rate. The circulating current is obtained either by galvanic anodes (also called

sacrificial anodes) or by an impressed current system [27].

Fig. 2.14. (a) CP by galvanic anode and (b) CP by impressed currents [27].

In the first case, CP is achieved through the galvanic coupling with a more active metal, as

shown in Fig. 2.14a. Aluminium and zinc are used for steel protection in sea water while

magnesium is employed in soil and fresh water [76]. The impressed current system makes use of

(a) (b)

31

a DC feeder (Figure 2.14b), with the positive pole connecting to the anode, generally an insoluble

metal (such as graphite or activated titanium) and the negative pole connected to the protected

structure. Galvanic anodes are typically used in high conductive environments (such as sea water)

and when a low protection current is required. Impressed current systems are usually adopted in

high resistive environments, generally soils and concrete, and are preferred when extended

structures must be protected, due to the higher flexibility in the current supply [77].

In practice, the impressed current CP is more effective to mitigate AC corrosion of buried

pipelines due to its flexible voltage and current outputs [78, 79], where the cathodic current is

impressed on pipelines through a DC power source by two mechanisms. The first one is that the

voltage or current of the output terminals of the DC power source keeps constant, while the

external resistance of the CP circuit varies to affect the CP potential [17]. For example, the soil

resistivity changes during dry-wet cycles of the soil, and the DC power source needs to adjust in

order to maintain the protection potential value. The other mechanism is that constant potential or

current from the output terminals is controlled by maintaining the pre-set pipe-to-soil potential vs.

a reference electrode. If this potential changes with environmental conditions, the output current

either increases or decreases. The first type is termed as ‘internal feedback’ CP system, where the

DC power source unit maintains a constant output of voltage/current by an internal stabilization

circuit. The desired CP potential on pipelines can be achieved by manual or semi-automatic

adjusting only, as shown in Fig. 2.15. In comparison, the second type of CP system can monitor

the CP potential in real time and maintain the value automatically by an ‘external feedback’ circuit

through an on-site reference electrode. However, the ‘external feedback’ CP system is more

complex in practical application for a long-distance transmission pipeline. While the ‘internal

feedback’ CP system is economic and convenient for installation in the field.

32

Fig. 2.15. Schematic diagram of the 'internal feedback' and 'external feedback' CP system

[17].

2.4.2 Ineffectiveness of CP by AC

Historically, it was thought that AC corrosion can occur on a cathodically protected pipeline

only when the peak of the positive part of AC wave form is more positive than the protection

potential [80]. Thus, a prevailing method to prevent AC corrosion is that applying cathodic

protection in accordance with industry standards could adequately control the AC-enhanced

corrosion [81]. However, multiple failures of pipelines under CP have been attributed to AC

corrosion [82-84]. It has been noted that the ‘conventional’ NACE recommended CP criterion of

-0.85 V Cu/CuSO4 electrode (CCS) is not adequate to protect buried pipelines in the presence of

AC interference [85]. For example, the 1986 German investigation of an AC corrosion failure

reported a high pitting rate despite CP current density of 1.5 to 2.0 A/m2 and on-potentials of –1.8

to –2.0 V (CCS) were applied on the buried pipeline [26]. Song et al. [86] discovered that

33

cathodically protected coupon at the cathodic potential of -1.0 V (SCE) still showed significant

corrosion rate and it increased with the higher of AC current density. Fu and Cheng [87] also

demonstrated that, in the absence of AC interference or at a low AC current density, e.g., less than

20 A/m2, a CP potential of -0.95 V (CCS) is able to provide a full protection over the steel.

However, when the AC current density is higher than 20 A/m2, the CP potential should be shifted

negatively.

New CP criteria based on field and laboratory studies as well as literature researches have been

proposed for full protection of pipelines from AC corrosion [88]. Fig. 2.16 shows the graphic

illustration of the CP criteria for buried steel pipelines. It is seen that AC corrosion could be

mitigated to a negligible corrosion rate if AC current density is below 70 A/m2 and the protection

current density is below 20 A/m2.

Fig. 2.16. New CP criterion based on AC and DC current densities [88].

34

Carpentiers et al. [89] stated that steel corrosion in CP condition can occur under the influence

of AC interference in the potential ranges that would passivate the steel if AC is absent. Due to the

local alkalization at the metal surface in CP condition, they recommended the protection potential

to stay below -1.15 V (SCE) at all times. At the same time, a new CP level in the range from -1.0

to -1.2 V (CCS) is considered effective and no overprotection conditions were established based

on laboratory results [90].

Fu and Cheng [87] proposed a new CP design criterion to protect pipelines from corrosion

under AC interference. Fig. 2.17 and Table 2.1 show the newly CP design with consideration of

AC interference for pipeline steel in carbonate/bicarbonate solution. It is seen that pipeline steel

can be effectively protected by the 100 mV cathodic polarization relative to the NACE

recommended criterion, when the AC current density is smaller than 20 A/m2. For AC current

densities between 20 and 500 A/m2, the CP potential should be shifted negatively to protect the

steel. The potential along the line ABCD refers to the ideal potential for protection of pipeline

from AC induced corrosion, where the suggested CP potential would provide a full protection to

the steel from AC corrosion. Area between lines ABCD and EFGH is the protection region and

the area left to the line ABCD is corrosion region. However, the possible detrimental effect should

be considered if the AC current density is very high. Area right to the line EFGH, is the

overprotective region. With consideration of the potential detrimental effect by more negative

potentials to pipelines, such as hydrogen induced cracking and cathodic disbondment of coating,

it is not recommended to use CP technique to protect the coated pipeline if the stray AC current

density is over 500 A/m2.

35

Fig. 2.17. New CP criteria with consideration of AC interference for pipeline steel in

carbonate/bicarbonate solution [87].

Table 2.1. New CP criteria with consideration of AC current interference for protection of

buried pipelines [87].

AC current density

(A/m2)

Level of AC

current density

Protection potential

(mV vs CCS)

Potential

difference (mV)

0 Zero -950 -100

20 Low -950 -100

100 Moderate -1050 -200

500 High -1150 -300

36

The AC interference could deviate the potential of the protected pipeline from the applied CP

value [91]. For example, Camitz et al. [92] studied the AC corrosion on steel coupons under

standard CP and various AC voltages in a long-term field investigation. Two series of tests were

performed: one at 10 V AC for almost 2 years, and the other at 30 V AC for 1.5 years. It was found

that the potential varied according to the applied AC voltage. During the positive half cycle, the

off-potential shifted in the anodic direction to values less negative than the applied CP value,

indicating that CP was periodically lost due to AC interference. In several cases, the potential

shifted to values even less negative than free corrosion potential. Besides, the characteristic

morphology of AC corrosion with large point-shaped attacks evenly distributed, and a few large,

deep local attacks were found across the surface.

It was confirmed by Xu et al. [17] that the AC would shift CP potential from the designed

value and the effect of AC on the CP performance depends on the cathodic potential applied on

the steel. At CP potential of -0.85 V (SCE), the AC application shifts the DC potential of the pipe

steel negatively, increasing the steel corrosion. When the CP potential is -1.0 V (SCE), the DC

potential of the steel is positively shifted, but the AC application would not be able to affect the

steel corrosion at a detectable level.

37

Chapter Three: AC Corrosion at Coating Defect on Pipelines1

3.1 Introduction

Generally, AC corrosion is not an issue when pipeline coatings are highly intact where a purely

capacitive behavior is shown. When the coating contains macroscopic damage or microscopic

defects (i.e., pinholes), the steel at the defect base can suffer from serious AC corrosion, even with

the application of CP [93, 94].

There have been extensive researches studying the AC-induced corrosion of carbon steel

pipelines experimentally and theoretically [95-100]. Coating performance plays a critical role in

the occurrence of AC corrosion of pipelines. Defects in coating can be introduced during

manufacturing and transportation, as well as pipeline construction. The geometry of a defect, such

as its size and shape, significantly affects the AC corrosion [101]. However, there has been so far

limited work to study the AC induced corrosion of the buried pipeline in simulated solutions under

coating defects. Xu et al. [102] studied the coated pipeline steel with different experimental

conditions under AC corrosion in a simulated soil solution, and they found that the AC interference

may cause destructive pitting on oil and gas transmission pipelines. Jiang et al. [103] simulated

corrosion occurring at coating defects under AC interference, and demonstrated that the defect size

affects remarkably the AC corrosion. Moreover, in preliminary investigation regarding the effect

of coating defects on pipeline under AC interference, Fu and Cheng [104] found that there was a

higher electrochemical dissolution activity of the coated steel electrode containing a 1 mm defect

1 This work has been published in Corrosion (D. Kuang, Y.F. Cheng, AC corrosion at coating defect on pipelines,

Corrosion 71, 2015, 267-276).

38

than that containing a 10 mm defect in the chloride-containing high pH solution. It was thus shown

that small defects induce a higher AC corrosion risk.

In this work, the effect of the geometry of coating defects, including the defect size and shape,

on AC corrosion of pipeline steel was studied in a deoxygenated, near-neutral pH bicarbonate

solution by direct current (DC) voltage analysis, potentiodynamic polarization curve

measurements and surface characterization. The essential role of the defect geometry in AC

corrosion was discussed, and the mechanistic aspect of localized corrosion occurring at the base

of the defect was analyzed.

3.2 Experimental

3.2.1 Specimen and solution

The coated steel specimens used in this work were fabricated from a sheet of X65 pipeline

steel coated with a high performance composite coating (HPCC) supplied by Bredero Shaw. The

chemical composition of X65 steel is (wt.%): C 0.04, Si 0.2, Mn 1.5, P 0.011, S 0.003 and Mo

0.02 and Fe balance. The HPCC is a multi-component, single layer coating system consisting of a

fusion bonded epoxy (FBE) primer, a polyethylene outer layer and a tie layer composed of a

chemically modified polyethylene adhesive, with an average thickness of 1.1 mm. Each specimen,

with a dimension of 50 mm × 50 mm × 10 mm, contained an artificial hole in its center with a

diameter of 5 mm, 7 mm, 10 mm and 20 mm, respectively. Additional specimens were prepared

with holes shaped of square and triangle in order to investigate the effect of the defect shape on

AC corrosion.

A dilute sodium bicarbonate (NaHCO3) solution with a pH of 7.5 was used to simulate the

groundwater. The solution contained 0.01 M NaHCO3 and purged with a 5% CO2/N2 mixture gas

39

prior to and throughout testing. The solution was made from analytic grade reagents (Fisher

Scientific) and ultra-pure water.

All tests were conducted at room temperature of 23 .

3.2.2 AC corrosion testing and DC potential derivation

The experimental setup for AC corrosion testing was shown in Fig. 3.1. The coated steel

specimen was used as working electrode (WE), a carbon rod was used as counter electrode (CE),

and a saturated calomel electrode (SCE) was used as reference electrode (RE). The distance

between WE and RE was about 1 mm in order to reduce the ohmic drop in potential measurements.

Fig. 3.1. Schematic diagram of the experimental setup for AC corrosion in the simulated

soil solution.

40

The AC was applied on the WE through a capacitor (50 V, 1000 μF) and a resistor (10 Ω, 0.5

W). The capacitor was used to block the DC signal, and the inductor prevented the flow of AC

current into the Solartron 1280C electrochemical measurement system. A data acquisition (DAQ)

system with an internal resistance of 144 kΩ was used to measure and analyze the potential

between the WE and the RE. Real-time DC potentials were derived through the DAQ, where an

home-developed software was used to separate the DC and AC voltages from the recorded mixed

voltage signal [99].

After a steady-state corrosion potential was achieved, potentiodynamic polarization curves

were measured under various applied AC currents at a potential scanning rate of 0.5 mV/s. To

ensure reproducibility of the testing data, each test was performed at least three times.

3.2.3 Surface characterization

The morphological features of the steel specimens were observed by an optical microscope

after AC corrosion testing.

3.3 Results

3.3.1 AC corrosion testing on coated steel electrodes containing a defect with varied sizes

Fig. 3.2 shows the DAQ-separated DC potentials of the coated steel electrode containing a

circular defect with 20 mm in diameter in the test solution. There are three potential regions: (a)

from 0 to 1200 s of testing, there is no AC applied on the electrode. The recorded DC potential is

corrosion potential; (b) from 1200 s to 2400 s, the AC-ON potential is recorded, and (c) after 2400

s, the AC application is stopped and the AC-OFF potential is recorded. It is seen that the DC

potential shifts negatively with the increase of AC current from 0 mA to 20 mA. Moreover, the

41

recorded DC potential profile contains more fluctuations at increasing AC currents. When the AC

is stopped at the 2400th s, the DC potentials cannot return to their original values. The new steady-

state DC potentials are less negative than the values recorded in the period of 0 ~ 1200 s.

0 400 800 1200 1600 2000 2400 2800 3200 3600 4000

-0.80

-0.78

-0.76

-0.74

-0.72

-0.70

-0.68

-0.66

-0.64

-0.62

-0.60

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

0 mA

1 mA

10 mA

20 mA


diameter defect at various AC currents in the test solution.

Fig. 3.3 shows the polarization curves measured on the coated steel electrode containing a 20

mm defect under various AC currents in the solution. The corrosion potential is shifted negatively

with the increasing AC current, which is consistent with the DC voltage measurements in Fig. 3.2.

Both anodic and cathodic current densities increase with the AC current, although it is less apparent

in the cathodic curves than it is in the anodic.

42

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0 -2.5-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

0 mA

1 mA

10 mA

20 mA


defect under various AC currents in the solution.

Fig. 3.4 shows the optical views of the coated steel electrode with a 20 mm defect after 12 h

of testing in the solutions under various AC currents. In the absence of AC, the electrode surface

is smooth, and does not show obvious corrosion signs, as shown in Fig. 3.4a. Upon the application

of AC, corrosion becomes apparent on the steel surface. Generally, the steel experiences uniform

corrosion, and the applied AC does not obviously affect the corrosion morphology. Figs. 3.5

through 3.7 show the DC potential, polarization curves, and corrosion morphology, respectively,

of the coated steel electrode containing a 10 mm diameter defect. Fig. 3.5 shows that the

dependence of the DC potential of the steel is identical to that observed on the electrode containing

a 20 mm defect in Fig. 3.2. The DC potential is shifted negatively under AC, and becomes more

negative with increasing AC. When AC is stopped, the AC-OFF potential is more negative than

43

the original DC potential. It can be discovered in Fig. 3.6 that the AC dependence of the measured

polarization curves of the steel electrode containing a 10 mm defect is also identical to that

observed on the electrode with a 20 mm defect in Fig. 3.3. The corrosion potential becomes more

negative with increasing AC. The anodic current density increases with AC, while the cathodic

current density does not show a good dependence.

Fig. 3.4. Optical views of the coated steel electrode containing a 20 mm defect after 12 h of

testing in the test solution at various AC currents (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20

mA.

44

0 500 1000 1500 2000 2500 3000 3500 4000-0.86

-0.84

-0.82

-0.80

-0.78

-0.76

-0.74

-0.72

-0.70

-0.68

-0.66

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

0

1 mA

10 mA

20 mA





-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0 -2.5

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

0 mA

1 mA

10 mA

20 mA

45

In the absence of AC, the steel at the 10 mm defect does not show apparent corrosion (Fig.

3.7a). Under 1 mA of AC, uniform corrosion is observed, as shown in Fig. 3.7b. When the AC is

increased to 10 mA and 20 mA, pitting corrosion occurs on the steel surface, as shown in Figs.

3.7c and d. Moreover, the size and depth of the pits increase with the AC.


solution at various AC currents after 12 h of test (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20 mA.

46

Figs. 3.8 through 3.10 show the DC potential, polarization curves, and corrosion morphology,

respectively, of the coated steel electrode containing a 5 mm diameter defect. As can be seen in

Fig. 3.8, the DC potential is shifted negatively upon AC application, and the value becomes more

negative under increasing AC. Moreover, the steady-state DC potential at individual AC is more

negative than that measured on the electrode containing the 10 mm defect. The potential

fluctuation is also more apparent.

0 400 800 1200 1600 2000 2400 2800 3200 3600-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

0 mA

1 mA

10 mA

20 mA



The polarization behavior, as a function of AC measured on the electrode containing the 5 mm

defect, is identical to that measured on the electrode containing a 10 mm defect in Fig. 3.6. The

47

corrosion potential is shifted more negatively and the anodic current density increases under the

rising of AC. The cathodic current density does not show a good dependence with AC.

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0 -2.5-1.3

-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

0 mA

1 mA

10 mA

20 mA



The corrosion morphology of the steel at the 5 mm defect shows that pitting corrosion occurs

under AC as low as 1 mA. At 10 mA or 20 mA, pitting corrosion on the steel becomes more

pronounced. In particular, pit depth and size increase remarkably at 20 mA of AC, as shown in

Fig. 3.10d.

48


solution at various AC currents after 12 h of test (a) 0 mA; (b) 1 mA; (c) 10 mA; (d) 20 mA.

3.3.2 AC corrosion testing on coated steel electrodes containing defects of varied shapes

Fig. 3.11 shows the DC potential of a coated steel electrode containing either a circular, square,

or triangular defect with an area of 78.5 mm2, i.e., the area of a 10 mm diameter defect, under 500

A/m2 AC current density in the test solution. The DC potential of the electrode is very similar for

each defect shape. Upon application of AC, the DC potential shifts negatively. When AC is

49

stopped, the DC potential reaches a relatively steady-state value, which is less negative than the

original potential prior to the application of AC.

0 500 1000 1500 2000 2500 3000 3500 4000-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

Time(S)

DC

Po

ten

tia

l (V

,SC

E)

circle

suqare

triangle



density in the test solution.

Fig. 3.12 shows the polarization curves measured on electrodes containing defects with the

three shapes under 500 A/m2 AC current density in the solution. There is little difference between

the measured curves.

50

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

circle

square

triangle


and triangular defect under 500 A/m2 AC current density in the test solution.

Figs. 3.13 and 3.14 show the DC potential and the polarization curves of the coated steel

electrodes containing either a circular, square, or triangular defect, each with an area of 38.5 mm2,

i.e., the area of a 7 mm diameter defect, under 500 A/m2 of AC current density in the test solution.

The DC potential was seen to vary, depending on the shape of the defect, with the DC potential of

the circular defect being the most negative, followed by that of the square, and then the triangular

defect. Although the polarization curves for each defect shape are not very different from each

other, the corrosion potential decreases, with the most negative being the circular, followed by the

square, and finally the triangular defect.

51

0 500 1000 1500 2000 2500 3000 3500 4000-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

circle

square

triangle




-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

circle

square

triangle


and triangular defect, each with an area of 38.5 cm2 under 500 A/m2 AC current density in

the test solution.

52

The measured DC potential and polarization curve are shown in Figs. 3.15 and 3.16,

respectively, for the defect area reduced to 19.6 mm2, which is equivalent to the area of a 5 mm

diameter defect. The differences in both DC potential and polarization curve are obvious. The

circular and triangular defects are associated with the most and the least negative DC potentials,

respectively. The circular defect creates the greatest anodic current density, followed by the square

defect; the triangular defect is associated with the smallest anodic current density.

0 500 1000 1500 2000 2500 3000 3500 4000-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

circle

square

triangle




53

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

circle

suqare

triangle


and triangular defect, each with an area of 19.6 cm2 under 500 A/m2 AC current density in

the test solution.

3.4 Discussion

3.4.1 Effect of defect size on AC corrosion of coated steel

Generally, the presence of defects in pipeline coating is inevitable. This work demonstrates

that the size of the defect is critical to AC corrosion occurring at the defect base. During the

corrosion of pipeline steel in CO2-purging, near-neutral pH bicarbonate solution, the anodic and

cathodic reactions include Fe oxidation and the reduction of bicarbonate ions, respectively [105]:

Fe → Fe2+ + 2e (3.1)

HCO3–

+ e → CO32– + H (3.2)

54

At small defects, the corrosion product Fe2+ does not diffuse freely out of the defect, due to

geometrical limitations. This would cause a local saturation of Fe2+ ions and CO32– ions to exceed

the solubility of iron carbonate (FeCO3), favoring the formation of FeCO3 scale:

Fe2+ + CO32– → FeCO3 (3.3)

While the iron carbonate scale formed at ambient temperature is not protective, the deposit

inside the defect generates a blocking effect that protects the steel from further corrosion. However,

at large defects, the generated Fe2+ ions could diffuse out of the defect and would not contribute

to a blocking effect. The resulting corrosion behavior is demonstrated in the measured polarization

curves on the steel electrode containing differently sized defects in Fig. 3.17, where the AC is not

applied. It is seen that, with a decrease in defect size, both anodic and cathodic current density

become smaller. This can be attributed to the blocking effect at the defect.

0 500 1000 1500 2000 2500 3000 3500 4000-0.700

-0.695

-0.690

-0.685

-0.680

-0.675

-0.670

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

20mm

10mm

5mm

DC potential

55

-8.0 -7.5 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

20 mm

10 mm

5 mm

Fig. 3.17. DC potential and polarization curves measured on the steel electrode where no

AC is applied, containing defects with different sizes in the test solution.

The application of AC shifts the DC potential of the steel negatively, no matter the size of the

defect. Fig. 3.18 shows the DC potential measured on the steel electrode containing differently

sized defects. It is seen that, at individual AC, with a decrease in defect size, the DC potential is

more negative. Moreover, with the increase of AC, the DC potential is more negative at a given

defect size. For example, for a diameter of 5 mm, the steady-state DC potential is about -0.75, -

0.85, and -0.90 VSCE at an AC of 1, 10, and 20 mA, respectively.

56

0 500 1000 1500 2000 2500 3000 3500 4000-0.76

-0.75

-0.74

-0.73

-0.72

-0.71

-0.70

-0.69

-0.68

-0.67

-0.66

-0.65

-0.64

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

20mm

10mm

5mm

(a)

0 500 1000 1500 2000 2500 3000 3500 4000

-0.88

-0.86

-0.84

-0.82

-0.80

-0.78

-0.76

-0.74

-0.72

-0.70

-0.68

-0.66

-0.64

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

20mm

10mm

5mm

(b)

57

0 500 1000 1500 2000 2500 3000 3500 4000-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

DC

Po

ten

tia

l (V

,SC

E)

Time(S)

20mm

10mm

5mm

(c)

Fig. 3.18. DC potential of the steel electrode containing differently sized defects at AC

currents of (a) 1 mA; (b) 10 mA; and (c) 20 mA.

The negative shift of corrosion potential with the defect size is also seen from polarization

curves in Fig. 3.19. Moreover, both anodic and cathodic current densities decrease with decreasing

defect size. Furthermore, as shown in optical views of corrosion morphology in Figs 3.4, 3.7 and

3.10, as defect size decreases, the threshold AC to induce pitting corrosion also decreases, i.e., at

a certain AC, there is a higher possibility for pitting corrosion to occur on coated steel containing

a smaller defect.

58

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2(a)

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

20 mm

10 mm

5 mm

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

20 mm

10 mm

5 mm

(b)

59

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0-1.2

-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

Po

ten

tia

l (V

,SC

E)

logi (A/cm2)

20 mm

10 mm

5 mm

(c)

Fig. 3.19. Polarization curves measured on the steel electrode containing differently sized

defects at AC currents of (a) 1 mA; (b) 10 mA; and (c) 20 mA.

Generally, upon application of AC, the AC would pass through the coating defect, due to the

local high conductivity. At a given AC, a small defect causes a high AC current density. It has

been demonstrated [49] that AC current density (iAC), rather than AC current or AC voltage, is

critical to corrosion induced by AC interference. With the decrease of the defect size, the iAC at the

defect increases, resulting in an enhanced corrosion and local anodic dissolution. Thus, the DC

potential decreases when the defect size becomes smaller. When the AC current density at defect

reaches a certain value, pitting corrosion occurs. This is attributed to accelerated local dissolution

at the defect base under high AC current density. The threshold value of iAC to initiate pitting on

pipeline steel in the near-neutral pH bicarbonate solution will be determined in the Chapter Four.

60

Furthermore, as can be seen from the polarization curves, with the decrease in defect size, both

anodic and cathodic current densities decrease, which is attributed to the blocking effect of the

scale accumulating at the defect. With the increase of iAC, localized corrosion at the defect is

accelerated, resulting in the generation and deposit of corrosion product, due to the geometrical

factor. Consequently, both anodic and cathodic reactions are inhibited.

3.4.2 Effect of defect shape on AC corrosion of coated steel

In this work, it is demonstrated that the shape of a coating defect affects AC corrosion of steel

only when the defect is small, such as 19.6 mm2 in area (Figs. 3.15 and 3.16), while this effect is

undetectable at large defects, such as 78.5 mm2 in area (Figs. 3.11 and 3.12). At large defects, the

local solution chemistry is identical or nearly identical to that of bulk solution. The exposed steel

behaves like a macroscopic electrode in the solution. Thus, the AC enhanced corrosion occurs

independently of the defect shape. However, at small defects, the mass transfer step and deposit of

a corrosion product become important. In particular, the differently shaped defects would have

different abilities to trap corrosion products due to geometrical effect, resulting in the dependence

of AC corrosion on defect shape.

The present work further shows that the circular and triangular defects are associated with the

most and least negative DC potentials, and the largest and smallest anodic current densities,

respectively. As discussed, with accelerated corrosion by AC, corrosion product deposits occur at

small defects due to limited geometry for mass transfer. At the sharp corners of a triangular defect,

the geometrical space is too small for a corrosion product to diffuse away. Thus, it is easy for the

corrosion product to accumulate. By contrast, it is relatively difficult for the corrosion product to

accumulate at a circular defect due to its open geometry. At the same time, the AC current density

61

at a triangular defect’s sharp corner is higher than that at a circular defect’s round edge.

Consequently, localized corrosion at a triangular defect generates more corrosion product to

deposit locally. Fig. 3.20 shows the morphology of the steel at defects with 19.6 mm2 area under

500 A/m2 AC current density in the test solution. It is seen that, due to the deposit of thick corrosion

products at the sharp corners of the triangular defect, the color at the corner is different from the

color at the center, where corrosion is actively occurring. This is also seen at the corners of the

square defect, where the color of the corners differs from that of the defect center. Thus, a lower

anodic current density is obtained on the triangular defect, and the higher anodic current density

is measured on the circular defect.

Fig. 3.20. Morphology of steel at defect with 19.6 mm2 area under 500 A/m2 AC current


3.5 Summary

The size of the coating defect is critical to AC corrosion of pipeline steel at the defect base. As

the defect size decreases, the threshold AC to induce pitting corrosion decreases, which is due to

1 mm

62

large AC current density generated at the small defect. Moreover, with the decrease in defect size,

both anodic and cathodic current densities decrease, which is attributed to the blocking effect of

corrosion product deposited inside the defect.

The geometrical shape of the coating defect affects AC corrosion when the defect is small (i.e.,

19.6 mm2 in area), while the effect is undetectable at large defects (i.e., 78.5 mm2 in area).

Furthermore, circular and triangular defects are associated with the most and least negative DC

potentials, and the largest and smallest anodic current densities, respectively. This is attributed to

the geometry of the triangular defect, the sharp corners of which favor the accumulation of a

corrosion product and limit its diffusion, thereby reducing corrosion.

63

Chapter Four: AC Induced Pitting Corrosion on Pipelines in Both High pH and Neutral

pH Carbonate/Bicarbonate Solutions2

4.1 Introduction

In addition to enhanced general corrosion on pipelines, the AC is able to result in localized

corrosion in the form of extensive corrosion pits [106]. It was found [49] that frequent oscillations

of passive current density occur in the presence of AC in a high pH carbonate/bicarbonate solution

where the pipeline steel is considered in passivity. The amplitude of the current oscillation

increases with the increasing AC current density. When the AC current density is up to 100 A/m2,

pits are formed on the steel surface. In a neutral pH bicarbonate solution, extensive corrosion pits

were also observed on the surface of steel electrodes when the AC current density is increased to

800 A/m2 [46]. However, the mechanism of AC-induced pitting corrosion has so far remained

unknown.

This work attempted to understand the AC-induced pitting corrosion on pipeline steel from the

viewpoint of the double-charge layer at the steel/solution interface in the presence of AC

application. Electrochemical measurements and surface characterization were conducted on the

steel in a soil-extracted, neutral pH bicarbonate solution and a high pH carbonate/bicarbonate

solution. The thresholds of AC current density to initiate corrosion pits on the steel were

2 This work has been published in Corrosion Science (D. Kuang, Y.F. Cheng, Understand the AC induced pitting

corrosion on pipelines in both high pH and neutral pH carbonate/bicarbonate solutions, Corrosion Science 85, 2014,

304-310.).

64

determined in both solutions. Mechanistic models were developed to illustrate the occurrence of

pitting corrosion on pipelines under AC interference in the different solution environments.

4.2 Experimental

4.2.1 Electrode and solution

Specimens used in this work were fabricated from a sheet of X65 pipeline steel, with a chemical

composition (wt.%): C 0.04, Si 0.2, Mn 1.5, P 0.011, S 0.003, Mo 0.02 and Fe balance. The steel

coupons were machined into 1 cm × 1 cm × 0.5 cm cubes, which were embedded in an epoxy

resin, leaving a working area of 1 cm2. The specimen preparation was controlled carefully to ensure

that there was no bubble and groove generating at the epoxy/steel interface. All electrodes were

subsequently ground with 800 grit and 1200 grit emery papers, and then cleaned in distilled water

and methanol.

A high pH carbonate/bicarbonate solution and a near neutral pH bicarbonate solution were used

to simulate two typical environments pipelines encountered in services. The high pH solution

contained 0.05 M Na2CO3 and 0.1 M NaHCO3, with a pH of 9.6. The neutral pH solution contained

0.01 M NaHCO3 purging with 5% CO2/N2 mixture gas, with a solution pH of 7.5. All solutions

were made from analytic grade reagents and ultra-pure water.

All tests were conducted at room temperature of 23 oC.

4.2.2 AC corrosion and electrochemical measurements

The real-time potential of the steel electrode was measured and analyzed under various AC

current densities through a home-developed AC/DC data acquisition (DAQ) system, which was

able to separate the DC and AC potential components from the recorded mixed voltage signal [46].

65

The measurement was performed through a National Instruments USB-6009 multifunction data

acquisition card, and a home-developed NI-DAQ data-logging software package based on the

LabVIEW program platform. The software package contained: (a) a data analysis module to

analyze the recorded AC/DC mixed signals obtained by the data acquisition card, and then process

the amplitude, frequency, phase and power spectrum calculation; (b) a data operation and display

interface to enter settings and show real-time potential-time plots; and (c) a data storage module

to save the analyzed data as a text file. This DAQ technique is capable of separating the DC and

AC voltage signals from the recorded potential on the specimen [46], where the steel electrode

was used as working electrode (WE), a carbon rod as counter electrode (CE) and a saturated

calomel electrode (SCE) as reference electrode (RE). The distance between WE and RE was 1 mm

in order to reduce the Ohmic drop in potential measurements. The steel electrode was ground and

polished to achieve an identical surface condition between tests at different AC current densities.

The AC current density was determined by a slide rheostat, and its value was obtained by

measuring the voltage on the resistor and divided by resistance and the electrode area. The AC

current density presented in this work was a root mean square (RMS) value. The capacitor was

used to block DC signals [46]. The frequency of the AC signals used in this test was 60 Hz.

When the open-circuit potential of the WE reached a relatively steady value, potentiodynamic

polarization curves were measured using a Solartron 1280C electrochemical system at a potential

scanning rate of 0.5 mV/s, starting from -1.0 V (SCE) and ended at 1.0 V (SCE).

4.2.3 Surface characterization

The surface morphology of the steel electrode after AC corrosion testing was observed by an

optical microscope.

66

4.3 Results

4.3.1 AC corrosion testing in high pH solution

Fig. 4.1 shows the time dependence of DC potential of X65 steel electrode at various AC

current densities in the high pH solution. Three potential regions are observed, i.e., (a) from 0 to

1200 s, there is no AC applied on the electrode. The recorded DC potential is corrosion potential

of the steel; (b) from 1200 s to 2400 s, the DC potential of the steel electrode under AC, i.e., AC-

ON potential, is recorded; and (c) after 2400 s, the AC application is stopped and the AC-OFF

potential is recorded.

0 500 1000 1500 2000 2500 3000 3500 4000-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

400 A/m2

500 A/m2

300 A/m2

200 A/m2

100 A/m2

DC

Po

ten

tia

l (V

,SC

E)

Time (s)

0 A/m2


AC current densities in the high pH solution: (a) corrosion potential before 1200 s; (b) AC-

ON potential from 1200 s to 2400 s; (c) AC-OFF potential after 2400 s.

67

It is seen that the corrosion potential of the steel reaches the value of -0.2 V (SCE) after 1200

s of immersion in the solution. Upon AC application, the DC potential is shifted positively at AC

current densities of 100 A/m2 and 200 A/m2. When AC current density is up to 300 A/m2, the DC

potential is shifted negatively. Moreover, with the increase of AC current density, the AC-ON

potential becomes more negative. When AC is stopped at the 2400th s, none of the DC potentials

can be back to their original value of -0.2 V (SCE). At 100A/m2 and 200 A/m2, the AC-OFF

potentials are less negative than the original corrosion potential; while at AC current density from

300 A/m2 to 500 A/m2, the AC-OFF potentials are much more negative than corrosion potential.

Fig. 4.2a shows the potentiodynamic polarization curve of the steel in the absence of AC in

high pH solution. It is seen that the steel can be passivated in the solution. Two oxidative current

peaks are observed at about -0.6 V (SCE) and -0.3 V (SCE), respectively, which is consistent with

the reported result in the same environment [107]. Fig. 4.2b shows the potentiodynamic

polarization curves of the steel at various AC current densities in the solution. It is seen that the

passive region becomes smaller compared to the curve in Fig. 4.2a. Moreover, the anodic current

density increases with the increasing AC current density. Fig. 4.3 shows the optical views of the

steel electrode after different testing times in the high pH solution under various AC current

densities. No pit is observed on the steel surface when AC current density is smaller than 200 A/m2

for the testing time up to 12 h. However, when AC current density is increased to 300 A/m2,

corrosion pits are observed on the electrode surface after only 5 min of testing. The size of pits

increases with time. This also applies for the AC current densities of 400 and 500 A/m2.

68

-7 -6 -5 -4 -3 -2

-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2

Cathodic

Pote

ntial (V

, S

CE

)

logi (A/cm2)

Anodic

(a)

-7 -6 -5 -4 -3 -2

-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2(b)

500 A/m2

400 A/m2 300 A/m

2200 A/m

2

Pote

ntial (V

, S

CE

)

logi (A/cm2)

100 A/m2


densities in high pH solution (a) 0 A/m2; (b) 100 - 500 A/m2.

69

Fig. 4.3. Optical views of the steel electrode after different testing times in the high pH


magnification bar is 500 μm for all photos.

0 A/m2

100 A/m2

200 A/m2

500 A/m2

400 A/m2

300 A/m2

70

4.3.2 AC corrosion testing in neutral pH solution

Fig. 4.4 shows the time dependence of DC potential of X65 steel electrode at various AC

current densities in the neutral pH solution. When AC is applied at the 1200th s, it can be seen that,

different from that observation in high pH solution, the DC potential is shifted negatively, with the

most negative value observed at 200 A/m2. As the AC is stopped at the 2400th s, the DC potential

jumps positively, and then tends towards the original corrosion potential. However, all AC-OFF

potentials are less negative than corrosion potential.

0 500 1000 1500 2000 2500 3000 3500 4000-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

500 A/m2

400 A/m2

300 A/m2

200 A/m2

100 A/m2

DC

Po

ten

tia

l (V

,SC

E)

Time (S)

0 A/m2


AC current densities in the neutral pH solution: (a) corrosion potential before 1200 s; (b)

AC-ON potential from 1200 s to 2400 s; (c) AC-OFF potential after 2400 s.

71

Fig. 4.5 shows the potentiodynamic polarization curves of X65 steel in the neutral pH solution

under various AC current densities. It is seen that the steel is in an active dissolution before and

after AC application. Moreover, the cathodic current density increases with the AC current density,

but the anodic current density almost copies each other. Fig. 4.6 shows the optical view of the steel

electrode after different testing times in the neutral pH solution. It is seen that there is no pit

generated even after 12 h of testing in the absence of AC. However, pits are formed under 100

A/m2 after 12 h of testing. With the increasing AC current density to 200 A/m2, pits are formed

after 5 min only. When the AC current density increases to 300 A/cm2 ~ 500 A/m2, the number

and size of pits rise apparently. Moreover, with the increasing of AC current density, the pits

become bigger and deeper, but the number decreases with the time increasing from 5 min to 12 h.

-7 -6 -5 -4 -3 -2-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2

Pote

ntia

l (V

, S

CE

)

logi (A/cm2)

0 A/m2

100 A/m2

200 A/m2

300 A/m2

400 A/m2

500 A/m2


densities in neutral pH solution.

72

Fig. 4.6. Optical views of the steel electrode after different testing times in the neutral pH


magnification bar is 500 μm for all photos.

0 A/m2

100 A/m2

200 A/m2

500 A/m2

400 A/m2

300 A/m2

73

4.4 Discussion

4.4.1 AC corrosion of pipeline steel in high pH solutions

The present work shows that the steel is in passivity in the high pH carbonate/bicarbonate

solution. As seen in Fig. 4.2, the anodic current peaks are associated with oxidation reactions at

individual potential. At -0.6 V (SCE), the observed peak is attributed to formation of FeCO3 and

Fe(OH)2 on the electrode surface [108-110]:

Fe + CO32- → FeCO3 + 2e (4.1)

Fe + HCO3- → FeCO3 + H+ + 2e (4.2)

Fe + 2H2O → Fe(OH)2 + 2H+ + 2e (4.3)

The second current peak observed at about -0.3 V (SCE) is due to the oxidation of ferrous

species to Fe2O3 and Fe3O4 [111-113]:

4Fe(OH)2 + O2 → 2Fe2O3 + 4H2O (4.4)

4FeCO3 + O2 + 4H2O → 2Fe2O3 + 4HCO3- + 4H+ (4.5)

6FeCO3 + O2 + 6H2O → 2Fe3O4 + 6HCO3- + 6H+ (4.6)

For the transitional zone in between, it is believed that two steps are involved, i.e., dissolution of

FeCO3 to complex Fe(CO3)22- and oxidation of ferrous ions to ferric substances.

Upon AC application, there is a remarkable change of the anodic polarization behaviour. While

the first current peak is still observed at about -0.6 V (SCE), the second peak is shifted positively

to about 0.3 V (SCE). After that there is no passive behaviour observed. Obviously, the applied

74

AC retards the formation of stable ferric oxide and de-passivates the steel. Moreover, with the

increase in AC current density, the anodic current density increases.

Under low AC current densities, such as those smaller than 200 A/m2, the AC induced

polarization is not sufficient to induce localized dissolution on the steel. With the increase of AC

current density to 300 A/m2, the dissolution current density is sufficiently large and small pits are

formed locally only after 5 min of testing. Thus, the AC current density of 300 A/m2 is considered

as the threshold of AC current density to initiate pits on steel in high pH solution. When the AC

current density is up to 400 and 500 A/m2, anodic dissolution current density further increases, and

pits form extensively on the electrode surface. Moreover, the number and size of pits increase with

time, showing that AC induced the further growth of pits. When AC is removed, the DC potential

cannot return to the original corrosion potential, indicating that the AC-induced activation effect

on the steel maintains even when AC is stopped. Thus, the AC effect on the steel is irreversible.

A model is developed to illustrate mechanistically the effect of AC on DC potential and

corrosion of steel, as shown in Fig. 4.7. Generally, charge-transfer reactions occur primarily in the

double-charge layer at electrode/solution interface due to its extremely strong electric field, which

can be disturbed by the applied AC. The steel is in passivity in the high pH solution, forming a

passive film with varied compositions of FeCO3, Fe(OH)2, Fe2O3 and Fe3O4. In the absence of AC,

the formed film can protect the substrate steel from further corrosion, as illustrated in Fig. 4.7a.

When the AC current density is 100 A/m2, the formed film is still effective to protect steel,

preventing cation ions, such as Fe2+, generated at the steel/film interface from flowing to the

solution, but contributing to the film growth (Fig. 4.7b). The double layer serves as a current

reservoir for the Faradic process. The extra positive charges accumulated in the double layer shift

positively the DC potential, as shown in Fig. 4.1. With the increasing AC current density to 200

75

A/m2, the anodic polarization during AC cycling generates more ferrous ions due to Fe oxidation,

and the cathodic polarization results in reduction of the passive film. Consequently, some cation

ions can leak out of the film towards the solution, as shown in Fig. 4.7c. The DC potential thus

becomes more negative compared with that measured at 100 A/m2. When the AC current density

increases to 300 A/m2, the enhanced cathodic polarization is able to break the passive film and

weaken the electric field strength of the double-charge layer, accelerating flowing of positive

charges out of the film. This results in less positive charges reserved and a decreased DC potential.

As shown in Fig. 4.7e, the higher the AC current density, the more breakage of the passive film

and thus the more negative of the DC potential. A further increase of AC current density to 500

A/m2 would not result in a DC potential that is much different from that measured at 400 A/m2,

which means that the positive charges remaining in the double-charge layer achieve a balance with

those releasing into the solution.

Fig. 4.7. Schematic diagram of the mechanistic model for AC corrosion of steel in high pH

solution, where the red box refers to the film.

Fe

(a)

100 A/m2

Fe

Fe

200 A/m2

Fe

Fe

Fe

(b)

(d) (e) (f)

Passive film

Ferrous ion 0 A/m2

(c) Leakage ferrous ion

300 A/m2 400 A/m

2 500 A/m

2

76

4.4.2 AC corrosion of steel in neutral pH solution

The steel shows active dissolution behaviour in the neutral pH solution, as shown in the

measured polarization curves in Fig. 4.5. The corrosion product, such as Fe(OH)2 and FeCO3, may

deposit on the steel surface, but would not have much protective ability due to its porous structure

[114].

Upon AC application, the DC potential is shifted negatively, and reaches the most negative

value at 200 A/m2. After AC is stopped, the DC potential reaches a steady state which is less

negative than the original corrosion potential, as shown in Fig. 4.4. At the same time, the

morphological observations in Fig. 4.6 show that the critical AC current density to initiate

corrosion pits on steel electrode in the neutral pH solution is 200 A/m2, below which general

corrosion occurs. Above the critical AC current density, pits can initiate only after 5 min of testing.

The pits become bigger and deeper with time. At individual time, the size and depth of the pit

increase with the AC current density.

A model described in Fig. 4.8 can be used to illustrate the effect of AC on corrosion of steel in

neutral pH solution. In the absence of AC, a layer of corrosion product deposits on the electrode

surface. When an AC current density of 100 A/m2 is applied, the AC enhanced electric field

accelerates diffusion of cation ions generated during corrosion of steel through the porous

corrosion product layer, as shown in Fig. 4.8b. Thus, the number of positive charges in the double-

charge layer decreases, resulting in a negative shift of DC potential in Fig. 4.4. When the AC

current density increases to 200 A/m2, more ferrous ions are generated and diffuse through the

corrosion product layer towards the solution, resulting in a further negative shift of corrosion

potential. At the same time, the corrosion product layer thickens due to the enhanced corrosion of

the steel. The length of the diffusive pathway of positive charges increases, reducing the leakage

77

of cation ions. Thus, with the further increase of AC current density, the corrosion product layer

thickens and the DC potential is shifted positively since more cation ions are stayed in the double-

charge layer. The AC current density of 200 A/m2 can be regarded as the critical current density

for pit generation.

Fig. 4.8. Schematic diagram of the mechanistic model for AC corrosion of steel in neutral

pH solution, where the red box refers to the corrosion product layer.

Generally, theories and models have been developed to illustrate pitting initiation on steels

which is in passivity. These include depassivation-repassivation theory [115], local acidification

theory [116], point defect models [117], and chemical-mechanical models [118, 119]. In these

classic theories, a common requirement for initiation of corrosion pits is the adsorption of

aggressive ions, such as chloride ions. This is different from the present work, where chloride ions

Fe

Fe

Fe

Fe

Fe

Fe

0 A/m2 100 A/m2 200 A/m2

(d) (e) (f)

300 A/m2 400 A/m2 500 A/m2

Ferrous ion

Corrosion product

(b) (c) Leakage ferrous ion (a)

78

are not present or at a quite low content. Moreover, carbon steel is not passivated in neutral pH,

diluted bicarbonate solution. Initiation of pitting corrosion on steel is primarily due to the AC

interference, as discussed.

4.5 Summary

Pitting corrosion can be induced on pipelines by enhanced AC current density in both high pH

and neutral pH solutions. Mechanistic models are proposed to illustrate the pitting initiation in the

presence of AC based on its effect on the passive film or corrosion product layer as well as the

charge distribution in double layer on the steel surface. In the high pH solution, the critical AC

current density to initiate pitting is approximately 300 A/m2, above which, the number and size of

pits increase with the AC current density and time. Generally, the passive film formed on the steel

surface serves as a barrier to keep the cation ions that are generated due to the AC anodic

polarization under small AC current densities. With the increase of AC current density, the passive

film is damaged due to AC induced cathodic polarization. The accelerated dissolution of steel

causes formation of pits locally.

In the neutral pH solution, the threshold AC current density to initiate pitting in the testing

system is approximately 200 A/m2. The corrosion product layer generated by active dissolution of

steel can be a barrier to prevent cation ions from flowing towards the solution. Due to the porous

structure, pits are generated at the pores. Generation of ferrous ions due to AC induced corrosion

thickens the corrosion product layer, resulting in decrease of the number of pits.

79

Chapter Five: Degradation of Epoxy Coatings on Pipelines in the Presence of AC

interference3

5.1 Introduction

Ideally, pipeline coatings provide an excellent barrier to separate the pipeline steel from the

soil environments. Epoxy coatings have widely been used for pipeline corrosion protection due to

their strong adhesion, high chemical resistance, and good processing characteristics [120]. Liquid

epoxy coatings are generally used in conjunction with fusion bonded epoxy (FBE) mainline

coating to protect the girth weld areas [121]. They can also be used in repair and rehabilitation of

existing pipelines [122].

Coatings suffer from frequent degradation and failure in various modes during service due to

a wide variety of reasons such as poor surface preparation, unqualified coating application, soil

stress, and mechanical damage by backfill [123]. Cathodic protection (CP) serves as a backup for

corrosion prevention. However, the CP potentially could cause cathodic disbonding (CD) of the

coating, especially at coating defects such as pinholes and holidays [124].

For pipelines adjacent to high voltage alternating current (HVAC) transmission lines, the AC

interference would affect pipeline corrosion at coating defects, as studied previously. It is expected

that the coating itself, particularly, its dielectric strength and permeability, can be affected by the

3 This work has been published in Corrosion'2016 annual conference (Da Kuang, Frank Cheng, Coating degradation

in the presence of alternating current interference, Corrosion’2016, paper no. 2016-7153, NACE, Vancouver, Canada,

Mar. 6-10, 2016.).

80

AC. To date, there has been limited work conducted to investigate the degradation of coating

properties and performance in the presence of AC interference.

In this work, three primary properties owned by pipeline coatings for corrosion protection, i.e.,

adhesion, water uptake and corrosion resistance, were measured on two coatings, i.e., FBE and

liquid epoxy, under various AC voltages. The morphology and structure of the coatings upon AC

application were characterized by optical microscope, scanning electron microscope (SEM) and

Fourier Transform-Infrared Spectroscopy (FT-IR). A conceptual model was developed to illustrate

the mechanistic aspect of the coating degradation induced by AC.

5.2 Experimental

5.2.1 Coatings, steel and solution

The FBE membrane and FBE coated steel specimens, and liquid epoxy membrane (200 μm in

thickness), supplied by Bredero Shaw and Canusa-CPS, respectively, were used in this work. The

coating membranes were cut into circular shape with a diameter of 5 cm.

The X65 pipe steel, with a chemical composition (wt%) C 0.04, Si 0.2, Mn 1.5, P 0.011, S

0.003, Mo 0.02 and Fe balance, was used for testing. The steel coupon was sealed by epoxy,

leaving a working area of 1 cm2. Prior to testing, the steel electrode was subsequently grounded

with 400, 800, 1000 and 1200 grit emery papers, and cleaned in distilled water and acetone.

The test solution, with a pH of 6.7, was an extracted soil solution, and its chemical composition

is shown in Table 5.1. The solution was made from analytic grade reagents and ultra-pure water

(18 MΩ∙cm in resistivity).

81

Table 5.1. Chemical composition of extracted soil solution (unit: mg/L)

Ca2+ Mg2+ Na+ K+ SO42- Cl- HCO3

-

20 15 17 3 45 26 104

5.2.2 Measurements of coating properties

The adhesion of coatings to steel substrate was evaluated by cathodic disbonding test. Fig. 5.1

shows the experimental setup. The coated steel specimen with 20 mm × 20 mm in area, which was

used as the working electrode, was connected to a DC power supply, where the steel was connected

to the negative pole and a platinum sheet was connected to the positive pole. A DC voltage of -3.5

V (saturated calomel electrode, SCE) was applied on the working electrode for 48 h. A carbon rod

was connected to AC signal source through a slide rheostat, and the AC current density flowing

between the working electrode and the carbon rod was varied from 0 to 500 A/m2 by adjusting the

AC voltage applied on the rheostat. A capacitor was used to block DC signal to flow into the AC

circuit. A 3.2 mm diameter holiday was created at the center of the coated specimen through the

coating to expose the steel substrate in a 3 wt% NaCl solution. The test cell was placed in a water

bath maintained at 65 . After testing, the coated steel specimen was taken out of the cell, and

cooled in air to ambient temperature. A utility knife was used to make eight radial cuts starting at

the holiday through coating to the steel substrate at 45° between the cuts. The coating was then

peeled off manually until a resistance was met against this force. The peel-off distances of the

eight directions were measured, and the average distance was defined as the CD radius.

82

Fig. 5.1. Schematic diagram of the experimental setup for CD test.

The water uptake of coatings was measured using the gravimetric cup method. A test cell was

designed to determine the water permeability in the FBE and liquid epoxy coating membranes.

The experimental setup is shown in Fig. 5.2. The membrane was mounted as a lid on a glass

container, which was filled with the soil solution. Various AC voltages from 0 to 50 V were applied

on the coating through a carbon rod installed in the glass container and a piece of conductive

adhesive was bonded on the coating membranes. The containers were kept in an oven at 65 with

a pre-equilibrated relative humidity of zero. The container with coating sealing was weighed

before test, and then weighed at a 24 h of interval. The obtained weight change was used to

calculate the water uptake.

83

Fig. 5.2. Experimental set-up for the test of permeability of coatings to water.

Electrochemical impedance spectroscopy (EIS) was measured on coated steel specimens

where were under various AC voltages. The EIS tests were conducted under a sinusoidal excitation

potential of 20 mV in the frequency range from 20 kHz to 1 mHz.

5.2.3 Structural and morphological characterization

The FT-IR can characterize the functional groups and cross-linking occurring in polymer

coatings. It is an accurate technique to determine the interactions between water molecules and the

thermoset structures at the molecular level. In this work, the FT-IR was used to characterize the

functional groups and structure of the epoxy coatings under various AC voltages. For all spectra

recorded, the coating samples experienced a 64-scan data accumulation in the range of 500-5500

cm-1 at a spectra resolution of 4.0 cm-1.

The cross-sectional morphology of the coatings after AC corrosion testing were observed using

a SEM.

84

5.3 Results

5.3.1 Coating disbonding tests

Fig. 5.3 shows the surface morphologies of FBE coated steel specimen after CD testing under

various AC current densities. It is seen that, generally, the coating disbonding area increases with

the increasing AC current density after CD test.

Fig. 5.3. Surface morphology of FBE coated steel specimen after cathodic disbonding test

under various AC current densities (a) 0, (b) 100 A/m2, (c) 200 A/m2, (d) 300 A/m2, (e) 400

A/m2, (f) 500 A/m2.

85

Fig. 5.4 shows the disbonding radius from the centre of the holiday as a function of AC current

density. It is seen that the CD radius of FBE coating is 3.37 mm in the absence of AC, which is

similar to that obtained previously [125]. Upon AC application, the CD radius increases obviously.

At the AC current density of 500 A/m2, the radius is up to 7.98 mm, more than twice of that

measured without AC. Thus, AC is able to promote coating disbondment.

3

4

5

6

7

8

500400300200100

dis

bondm

ent

radiu

s (

mm

)

AC current density (A/m2)

0

Fig. 5.4. Coating disbonding radius as a function of AC current density measured in FBE

coated steel specimens.

Fig. 5.5 shows the surface morphology of the liquid epoxy coated steel specimens after CD

testing under various AC current densities. It is seen that the disbonding area also increases with

the AC current density.

86

Fig. 5.5. Surface morphology of liquid epoxy coated specimens after CD test under various

AC current densities (a) 0, (b) 100 A/m2, (c) 200 A/m2, (d) 300 A/m2, (e) 400 A/m2, (f) 500

A/m2.

Fig. 5.6 shows the disbonding radius from the centre of the holiday as a function of AC current

density. The CD radius is 4.07 mm in the absence of AC. Upon AC application, the CD radius

increases. At the AC current density of 500 A/m2, the radius is up to 10.09 mm. Obviously, the

AC is also able to enhance cathodic disbondment of the liquid epoxy coating. As the CD tests are

conducted under the identical conditions on both coatings, the liquid epoxy coating has a larger

disbonding radius than FBE at individual AC current density. Although AC would increase the

disbonding of coatings, the FBE shows a higher CD resistance under AC interference.

87

4

5

6

7

8

9

10

500400300200100

dis

bondm

ent

radiu

s (

mm

)


0

Fig. 5.6. Coating disbonding radius as a function of AC current density measured in liquid

epoxy coated steel specimens.

5.3.2 Water upktake tests

Fig. 5.7 shows the weight-loss of the testing containers sealed by individual coating

membranes as a function of time under various AC voltages. It is seen that the total weight of the

container decreases with time, indicating the evaporation of water from the solution. Moreover,

the weight loss increases with the increasing AC voltages. Thus, AC could enhance the water

uptaking by both coatings. Compare the two coatings under testing, there is a larger weight loss

for liquid epoxy coating than FBE at individual AC voltage.

88

0 2 4 6 8 1052

54

56

58

60

62

64

66

68

70

72

We

igh

t (g

)

Time (d)

0 V

10 V

20 V

30 V

40 V

50 V

(a)

0 2 4 6 8 1052

54

56

58

60

62

64

66

68

70

72

0 V

10 V

20 V

30 V

40 V

50 V

Time (d)

We

igh

t (g

)

(b)

Fig. 5.7. Time dependence of the weight of the coating-sealed test container under various

AC voltages: (a) FBE coating; (b) liquid epoxy coating.

89

The water transmission rate (WTR) has been used to characterize quantitatively the water

permeability of polymeric coatings [126]. The WTR equals to the slope of the weight-loss vs. time

divided by the exposed area of the coating:

tA

GWTR

(5.1)

where ΔG is the weight-loss (g), t is time (h), and A is the exposed area of the coating (m2). The

WTR of the two coatings under various AC voltages are calculated and shown in Fig. 5.8. It is

seen that the WTR increases with the AC voltage. At individual AC voltage, the liquid epoxy

coating has a larger water permeation than FBE.

0 10 20 30 40 500

500

1000

1500

2000

2500

3000

3500

WT

R (

g/m

2d

)

AC voltage (V)

FBE

Liquid epoxy

Fig. 5.8. Water transmission rate of the coatings under various AC voltages.

90

5.3.3 EIS measurements

Figs. 5.9 and 5.10 show the Nyquist diagrams measured on the coated steel specimens after 5

and 30 days of immersion in the test solution under various AC voltages. In the absence of AC

interference, the Nyquist diagram on FBE is featured with a big, incomplete semicircle. This is the

typical capacitive behavior associated with a high coating resistance in excess of 109 Ω∙cm2. Upon

application of various AC voltages, depressed semi-circles are measured. With the increase of the

AC voltage, the size of the semi-circle decreases, suggesting the decreasing coating resistance

under AC interference. Compared with FBE, the liquid epoxy coating shows a smaller resistance.

The AC voltage also results in the decrease of the size of the semicircle. Moreover, the semicircle

is slightly smaller than that for FBE at individual AC voltage. After 30 days of immersion as shown

in Fig. 5.10, the size of the semicircle further decreases, indicating the decreasing barrier property

of the coatings. The coating resistances decrease to 107 Ω∙cm2 and 106 Ω∙cm2 for FBE and liquid

epoxy coating, respectively, in the absence of AC. Moreover, with the increase in AC voltage, the

semicircle is smaller. Similarly, the semicircle for liquid epoxy is slightly smaller than that for

FBE at individual AC voltage. Another important phenomenon is that one more semicircle is

observed in the low frequency range after 5 days of immersion for liquid epoxy coating (Fig. 5.9b)

and after 30 days of immersion for both coatings (Fig. 5.10). Generally, the high-frequency

semicircle is associated with the coating properties, and the low-frequency one is attributed to the

corrosion reaction at the steel/coating interface [127]. Thus, with the increase of the immersion

time, corrosion occurs on the steel under coating due to permeation of water and corrosive species.

Moreover, the steel coated with liquid epoxy starts to have two semicircles in Nyquist diagrams

after only 5 days of immersion, indicating its poorer corrosion resistance than FBE.

91

0 2 4 6 8 100

2

4

6

8

10

0 5 10 15 20 25 30 35 400

50

100

150

-Z''

(M

cm

2)

Z' (M cm2)

0 V

Z' (M cm2)

-Z''

(M

cm

2)

10 V

20 V

30 V

40 V

50 V

(a)

0 2 4 6 8 100

2

4

6

8

10

(b)

0 5 10 15 20 25 30 35 40 450

5

10

15

20

25

30

35

40

45

Z' (M cm2)

-Z''

(M

cm

2)

0 V

-Z''

(M

cm

2)

Z' (M cm2)

10 V

20 V

30 V

40 V

50 V

Fig. 5.9. Nyquist diagrams measured on coated steel specimens after 5 days of immersion

in the solution under various AC voltages: (a) FBE (b) liquid epoxy coating.

92

0 2 4 6 8 10 120

2

4

6

8

10

12

(a)

-Z''

(M

cm

2)

Z' (M cm2)

0 V

10 V

20 V

30 V

40 V

50 V

0 2 4 6 8 10 120

2

4

6

8

10

12

(b)

-Z''

(M

cm

2)

Z' (M cm2)

0 V

10 V

20 V

30 V

40 V

50 V

Fig. 5.10. Nyquist diagrams measured on coated steel specimens after 30 days of

immersion in the solution under various AC voltages: (a) FBE, (b) liquid epoxy coating.

93

5.4 Discussion

5.4.1 Effect of AC on coating performance

The present work demonstrates that AC is also able to enhance disbonding of epoxy coating

from the steel substrate. The applied AC could facilitate the solution alkalization, which induces

the adhesion loss of epoxy coatings to the substrate steel. The hydrogen evolution and the resulting

bubbling due to enhanced cathodic reaction in the presence of AC also contribute to the coating

disbonding. In this work, it has been demonstrated that AC could enhance more serious

disbondment of liquid epoxy coating from the steel substrate.

The coating performance is highly dependent upon its resistance to water permeation. In this

work, it is confirmed that AC can increase the water permeation through both FBE and liquid

epoxy coatings. As shown in Fig. 5.8, the WTR of FBE under 50 V of AC voltage is about 4 times

of the one in the absence of AC interference, while the WTR of liquid epoxy coating under the

same AC voltage is nearly 7 times more than the one without AC. Furthermore, the coating

resistance decreases under AC application, and the resistance of liquid epoxy coating drops more

rapidly than FBE. Obviously, AC would degrade the properties and performance of the two

coatings, as indicated by the increased disbondment (reduced adhesion), increased water uptake,

and decreased coating resistance.

5.4.2 Effect of AC on coating structure

To understand the degradation of coatings in the presence of AC interference, SEM was used

to characterize the structure of the coating. Figs. 5.11 and 5.12 show the cross-sectional view of

FBE and liquid epoxy coatings after 30 days of immersion in the solution in the absence and

presence of applied AC voltage of 50 V, respectively. It is seen from Fig. 5.11a that FBE contains

94

a compact structure, with some micropores. Upon application of the AC voltage, more micropores

with a larger size are generated in the coating, as seen in Fig. 5.11b. In Fig. 5.12a, the liquid epoxy

coating is featured with a loose structure, with microcracks present in the coating. Under the effect

of AC, the coating becomes looser, with some tortuous, deep microvoids observed (Fig. 5.12b).

Fig. 5.11. SEM images of the cross-sectional view of FBE coating after 30 days of

immersion: (a) without AC; (b) AC voltage of 50 V.

Fig. 5.12. SEM images of the cross-sectional view of liquid epoxy coating after 30 days of

immersion: (a) without AC; (b) AC voltage of 50 V.

95

Figs. 5.13 and 5.14 show the FT-IR spectra of FBE and liquid epoxy coatings under various

AC voltages after 30 days of immersion in the test solution. The characteristic functional groups

identified are listed in Table 5.2. It is seen that three characteristic absorption peaks are observed

in the spectrum of FBE coating. The peak at 831 cm-1 is attributed to C-O bond in the oxirane

group. The peak at 2873 cm-1 is the C-H tension of the methylene group. The third one is located

at 4623 cm-1, which corresponds to a combination bond of the second overtone of epoxy ring

stretching with the fundamental C-H at 2873 cm-1 [128]. As for the liquid epoxy coating, the peak

at 4623 cm-1 is not observed, as shown in Fig. 5.14, due to the introduction of inclusions, which

makes the structure of liquid epoxy different from the typical epoxy resin coatings.

6000 5400 4800 4200 3600 3000 2400 1800 1200 600

5215 cm-1

4623 cm-1

4066 cm-1

3600 cm-1 2873 cm

-1

1509 cm-1

1608 cm-1

831 cm-1

50 V

40 V

30 V

20 V

10 V

0 V

ab

sorb

an

ce

wavenumber (cm-1)

Fig. 5.13. FT-IR spectrum of FBE coating after 30 days of immersion in the test solution

under various AC voltages.

96

Table 5.2. Characteristic bands of epoxy coating in the FT-IR spectrum

Band (cm-1) Assignment

5215

4623

4066

3600

2873

1608

1509

831

Combination asymmetric stretching and bending of O-H [129, 130]

Overtone of C-H stretching of the aromatic ring [131]

Stretching C-H of aromatic ring [132]

O-H stretching [133]

C-H stretching [134]

Stretching C=C of aromatic rings [135]

Stretching C-C of aromatic rings [136]

Stretching C-O-C of oxirane group [136]

6000 5400 4800 4200 3600 3000 2400 1800 1200 600

50 V

40 V

30 V

20 V

10 V

0 V

4066 cm-13600 cm

-1

2873 cm-1

1509 cm-1

1608 cm-1

831 cm-1

ab

sorb

an

ce

wavenumber (cm-1)

Fig. 5.14. FT-IR spectrum of the liquid epoxy coating after 30 days of immersion in the test

solution under various AC voltages.

97

Moreover, two types of water bands are observed in the spectrum of FBE coating, i.e., highly

mobile free water molecules at about 3600 cm-1 and the combination of asymmetric stretching and

bending of hydroxyl vibrations located at 5215 cm-1 [129, 130]. This combined band is not

observed in the spectrum of liquid epoxy coating either due to its coating modification.

It can be seen from Fig. 5.13 that the bands at 3600 cm-1 and 5215 cm-1, representing the

absorbed water in FBE coating, are enhanced in their absorbance intensities with the increase of

AC voltage. This indicates that the water uptake is increased by AC application. This result

provides a direct evidence that the water permeation into FBE coating increases with the increasing

AC voltage. More importantly, the three absorption peaks located between 1800 cm-1 and 2400

cm-1 are not the characteristic functional groups of epoxy resin, but are observed under the AC

application. It is thus assumed that the structure of FBE coating can be changed under AC

interference. These peaks are related to C≡C stretching, indicating the degradation of the coating

in the presence of AC. For liquid epoxy coating, the band at 3600 cm-1, representing the absorbed

water, is increased in its intensity with the increasing AC voltage. Thus, AC facilitates the water

uptake in the coating. Moreover, similar to FBE, a new absorption peak is observed at 2400 cm-1,

which is associated with the C≡N functional group. The intensity of the peak increases with the

AC voltage. Thus, AC application changes the structure of the liquid epoxy coating, affecting the

coating performance.

A conceptual model is developed to illustrate the mechanistic aspect of the AC induced coating

degradation, as shown in Fig. 5.15. Generally, epoxy resin is a highly polar coating containing

hydroxyl groups. Water is also a polar molecule and is capable of forming hydrogen bonding with

the hydroxyl groups. Thus, the epoxy coatings are permeable to water. In the presence of AC, an

electric field is generated on the coating, which is able to polarize the functional groups, increase

98

the coating polarity and enhance the water permeation. Thus, a layer of water molecules and ions

from the solution is formed by bonding with the polar groups on the coating surface, which is

called the polar effect of the AC interference.

Fig. 5.15. Schematic diagram of the conceptual model for illustrating the coating

degradation in the presence of AC interference.

Moreover, it was proposed [46] that the AC application could generate a ‘vibrating’ effect due

to the low frequency alternating electric field exerting on the reactants in the solution. The mobility

of solvated ions in the solution is usually slow, and is thus sensitive to the low frequency electric

field force. Upon application of a low frequency AC of 60 Hz in this work, there is a sufficient

time for the ions to be accelerated by the electric field. Consequently, the kinetic energy of

polarized ions and water is increased, enhancing their permeation into the coating through

nanopores and microvoids. The accelerated ions, once entering the coating, could impact the

99

functional groups and change the epoxy structure, as shown in the FT-IR spectra where new

absorption peaks are observed in the presence of AC. With the increase of the AC voltage, the

enhanced AC electric field could polarize more functional groups in the coating to absorb water

and ions, resulting in coating degradation.

5.5 Summary

The AC enhances the disbondment of coatings from the steel substrate. Moreover, the AC

increases the water uptake by the coatings, and decreases the coating resistance. The coating

degradation is increased with the increasing AC voltage. Compared with FBE, the liquid epoxy

coating suffers from more degradation in the presence AC interference.

Under AC interference, the morphological characterization shows that the coatings become

more loose and porous. The structural measurements by FT-IR show that the AC application

results in both the enhanced absorbance intensities of characteristic functional groups contained in

the coatings and the appearance of new absorption peaks. Thus, the AC is able to result in structural

changes of the coatings.

According to the proposed conceptual model to understand the coating degradation in the

presence of AC interference, the AC can generate an electric field on the coating to polarize the

functional groups to enhance the absorption of water molecules and ions. Moreover, the vibrating

effect induced by AC could accelerate the entry of polarized ions into the coating, and degrade the

functional groups of the coatings.

100

Chapter Six: Probing Potential and Solution pH under Disbonded Coating on Pipelines in

the Absence of AC Interference4

6.1 Introduction

Disbondment of pipeline coatings can occur by a number of mechanisms [137]. In addition to

poor surface preparation and thermal cycling during pipeline operation, cathodic disbondment is

an important mechanism that results in lost adhesion, which usually starts at a holiday or pinhole.

Cathodic protection (CP) at coating faults could elevate the electrolyte pH at the holiday through

enhanced cathodic reduction of dissolved oxygen or water [138]. The alkalization of the local

solution can weaken the bond of the coating primer to the steel, causing coating disbondment.

The shielding effect of coating disbondment on CP penetration into the disbonding crevice has

been investigated [139-142]. Parametric effects, such as solution resistivity, size of the holiday,

temperature, CP potential, and disbonding geometry have been tested. It is acknowledged that CP

can be shielded from reaching the disbonded crevice bottom. In addition, there are numerous

environmental conditions that can affect the CP shielding behavior.

The steel under a disbonded coating, especially the disbondment bottom, can experience

corrosion by anodic dissolution, while the open holiday remains under effective CP. The separation

of anodic and cathodic reactions could facilitate the occurrence of localized corrosion, which has

frequently been observed on cathodically protected pipelines where the coating has disbonded.

4 This work has been published in Materials Performance (D. Kuang, Y.F. Cheng, Probing potential and solution pH

under disbonded coating on pipelines, Materials Performance 54, 2015, 40-45).

101

Fusion-bonded epoxy (FBE) is a commonly used pipeline coating that is compatible with CP

[123]. However, there has been limited work to investigate the shielding effect of FBE induced by

geometrical factors (i.e., disbonding from a holiday) on CP, as well as the time dependence of this

effect. This work attempted to determine the CP shielding behavior under disbonded FBE through

in-situ probing of local potential and solution pH distributions under the coating. The effects of

CP potential as well as the disbonding thickness and depth were determined.

6.2 Experimental


Steel coupons and coating used in this work were X65 pipeline steel and FBE, respectively.

The chemical composition of the steel (wt. %) is 0.04 C, 0.2 Si, 1.5 Mn, 0.011 P, 0.003 S, 0.02

Mo, and balance Fe. Prior to testing, the steel surface was ground with 120, 240, 400 and 800 grit

emery papers, followed by cleaning in distilled water and methanol.

A near-neutral pH (7.5) bicarbonate solution was used to simulate the electrolyte trapped under

the disbonded coating. The solution was 0.01 M sodium bicarbonate (NaHCO3), and was purged

with 5% carbon dioxide (CO2)/nitrogen (N2) for 48 h prior to the testing.

6.2.2 Simulated crevice cell

Fig. 6.1 shows a home-designed experimental setup to simulate the crevice generated by

coating disbondment. The dimension of the X65 steel plate was 200 × 25 × 20 mm. To prepare an

artificial disbondment, the FBE membrane (200 μm in thickness) was applied on the steel surface

using a double-sided self-adhesive tape. The gap between FBE and the steel was defined as the

disbonding thickness. Tape with a known thickness was layered to establish the desired gap, which

102

was verified with a coating thickness gauge. The FBE membrane was applied to the tape, which

was then applied on the steel surface. The tape was removed to form artificial FBE disbondment

over the steel. The boundaries of the steel/coating assembly were sealed with an epoxy resin.

Fig. 6.1. Schematic diagram of the experimental setup simulating a disbonding crevice

under coating and the potential/solution pH measurements.

A 10 mm diameter hole was opened on the coating to simulate a holiday, with six potential/pH

micro probes installed at distances of 30, 60, 90, 120, 150 and 180 mm from the holiday. The

distance of the probing position to the open holiday was defined as the disbonding depth.

The corrosion potential of the steel in the solution was measured as -0.755 V vs. saturated

calomel electrode (SCE). Various CP potentials were applied to the coated steel through a DC

power supply with a platinum foil as the counter electrode, using SCE as the reference electrode

103

and the steel as the working electrode. The local potential can be measured by SCE reference

electrode through the AC/DC real-time signal measurements and the local solution pH was

measured through an Oaklon Acorn pH 5 meter. All tests were conducted at 23 .

6.3 Results

6.3.1 Distribution of local potential under disbonded coating

Fig. 6.2 shows the time dependence of the distributions of local potential under disbonded

coating (disbonding thickness of 120 μm) at varied disbonding depths from the open holiday,

where the steel was either at corrosion potential or at CP potentials of -0.875 V (SCE) and -0.975

V (SCE), respectively. It is seen that, prior to CP application, the local potentials at all probing

positions are about -0.755 V (SCE), which is the corrosion potential of X65 steel in the test

solution. When the potential of -0.875 V (SCE) is applied, the potential at the holiday (i.e., 0 mm

in the figure) is the applied CP value. However, the potential at the position of 30 mm from the

holiday is less negative (i.e., -0.810 V (SCE) after 48 h of testing). With the increase in the

disbonding depth (i.e., the probing position is farther away from the holiday), the local potential is

less negative than that at the 30 mm position, but the potential difference is not distinguishable. At

the CP potential of -0.975 V (SCE), the potential at the holiday is still the applied value, but the

local potentials at the probing positions are shifted less negatively. With the increase in disbonding

depth, the potential becomes less negative. The disbonding depth at the local potential is not

distinguishable as this CP level is increase to 150 mm.

104

0 10 20 30 40 50-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

P

ote

ntia

l (V

, S

CE

)

Time (h)

(a)

0 10 20 30 40 50-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50

(b) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

Time (h)

Pote

ntial (V

, S

CE

)

105

0 10 20 30 40 50-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50

(c) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

Time (h)

Po

tentia

l (V

, S

CE

)

Fig. 6.2. Time dependence of the distributions of local potential under disbonded coating

(disbonding thickness of 120 μm) at varied disbonding depths from the open holiday where

the steel was either at (a) corrosion potential at (b) CP potential -0.875 V (SCE) and (c) -

0.975 V (SCE) respectively.

Thus, the applied CP can be shielded from reaching the coating disbondment. With the increase

in disbonding depth toward the disbondment bottom, the CP shielding is more apparent. The

shielding effect can be mitigated by the application of more negative CP potentials.

Fig. 6.3 shows the distributions of local potential under disbonded coating at varied disbonding

depths from the open holiday, where the CP potential of -0.875 V (SCE) is applied under various

disbonding thickness. Identical to previous results, the CP is shielded from reaching the

disbondment. Only at the open holiday, the measured value is the same as the applied CP potential.

Under the coating disbondment, the potential tends to be less negative.

106

0 10 20 30 40 50-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

(a)

Time (h)

P

ote

ntial (V

, S

CE

)

0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

0 10 20 30 40 50-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

Time (h)

(b) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

P

ote

ntia

l (V

, S

CE

)

107

0 10 20 30 40 50-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

Time (h)

(c) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

P

ote

ntial (V

, S

CE

)

Fig. 6.3. Distributions of local potential under disbonded coating at varied disbonding

depths from the open holiday where the CP potential of -0.875 V (SCE) is applied under

various disbonding thickness (a) 120 μm, (b) 240 μm, (c) 360 μm.

Moreover, with the increase in disbonding thickness, the CP shielding effect becomes less

significant. For example, at the dsibonding thickness of 120 μm, the local potential at the probing

position of 30 mm is about -0.810 V (SCE). When the disbonding thickness is increased to 240

and 360 μm, the potentials at the same position are -0.855 and -0.865 V (SCE), respectively.

Therefore, as the coating disbondment becomes wider (i.e., with an increased disbonding

thickness), the CP shielding effect is less significant.

108

6.3.2 Distribution of solution pH under disbonded coating

Fig. 6.4 shows the time dependence of the distributions of local solution pH under disbonded

coating (disbonding thickness of 120 μm) at various disbonding depths, where the steel was either

at corrosion potential or at CP potentials of -0.875 and -0.975 V (SCE), respectively. Prior to CP

application, the solution pH is ~7.5, the value of the prepared solution, at all probing positions.

Upon CP application, the solution pH is elevated. Moreover, when the CP potential is more

negative, the solution pH is further elevated at individual probing positions. For example, at the

open holiday, the steady-state solution pH is 8.5 at -0.875 V (SCE) and 11.0 at -0.975 V (SCE)

after 48 h of testing. However, the CP driven pH elevation becomes less obvious with the

increasing disbonding depth, especially at the disbondment bottom.

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

(a) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

Time (h)

pH

109

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

(b) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

Time (h)

pH

0 10 20 30 40 507.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

12.5

13.0

Time (h)

(c) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

pH

Fig. 6.4. Time dependence of the distributions of local solution pH under disbonded

coating (disbonding thickness of 120 μm) at varied disbonding depths from the open

holiday where the steel was either at (a) corrosion potential at (b) CP potential -0.875 V

(SCE) and (c) -0.975 V (SCE) respectively.

110

Fig. 6.5 shows the distributions of solution pH under disbonded coating at varied disbonding

depths from the open holiday where the CP potential of -0.875 V (SCE) is applied under various

disbonding thicknesses. The applied CP is able to elevate solution pH, especially at the open

holiday. With the increasing disbonding depth, the solution pH tends to be the value of the

originally prepared solution. As the disbonding thickness increases, the CP induced pH elevation

becomes more obvious, even at the disbondment bottom. For example, at the disbonding thickness

of 120 μm, the solution pH at the disbondment bottom (i.e., 180 mm from the holiday) is ~ 7.5.

This indicates that the CP does not penetrate into the disbondment bottom. When the disbonding

thickness is increased to 240 and 360 μm, the solution pH at the disbondment bottom is 7.9 and

8.1, respectively. Thus, as the coating disbondment become wider, the CP enhanced pH elevation

is more appreciable.

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

Time (h)

(a) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

pH

111

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

Time (h)

(b) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

pH

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

12.0

Time (h)

(c) 0 mm

30 mm

60 mm

90 mm

120 mm

150 mm

180 mm

pH

Fig. 6.5. Distributions of solution pH under disbonded coating at varied disbonding depths

from the open holiday where the CP potential of -0.875 V (SCE) is applied under various

disbonding thickness (a) 120 μm, (b) 240 μm, (c) 360 μm.

112

6.4 Discussion

6.4.1 Effect of applied potential on CP shielding

In deoxygenated neutral pH bicarbonate solutions, the anodic and cathodic reactions during

corrosion of pipeline steel are primarily the iron oxidation and reduction of water respectively. In

the absence of CP, the steel corrodes at both the holiday and under the disbonded coating. Upon

CP application, the cathodic reaction is enhanced, resulting in the generation of OH- to elevate the

solution pH. With the negative shift of CP potential, the enhancement is more apparent. Thus, the

solution pH is further elevated.

When CP is applied on a coated steel electrode containing a holiday, the CP is primarily applied

on the open holiday. The solution pH at the holiday is elevated with the negative shift of CP

potential, as shown in Fig. 6.4. The non-uniform distribution of solution pH from the open holiday

to the disbondment indicates that the CP induced pH elevation is not fully realized under the

disbonded coating.

The measurements of the potential distribution from the holiday toward the disbondment

bottom indicate that the applied CP is shielded, at least partially, under the disbonded coating, as

shown in Fig. 6.2. In addition, it can be found that the potentials are consistent with the pH values.

At the disbonding thickness of 120 μm, only the open holiday is under full CP potential. With the

increasing distance from the holiday (i.e., the increasing disbonding depth), the local potential

tends to be less negative until the steady state corrosion potential is reached. Due to the shielding

effect, the steel under the disbonded coating is not under fully cathodic protection. The potential

records also indicate that, in order to cathodically protect the steel under coating disbondment, the

CP potential must be sufficiently negative. However, hydrogen evolution must also be considered.

113

6.4.2 Effect of disbondment thickness on CP shielding

The shielding effect of coating disbondment on CP permeation is affected by the disbonding

thickness, as shown in Figs. 6.3 and 6.5. The measurements of both local potential and solution

pH under disbonded coating show that the CP shielding tends to be mitigated when the disbonding

thickness is increased, and the potentials under disbonded coating approach those at the open

holiday. Moreover, the pH elevation under disbondment is more apparent. Thus, the geometrical

factor of the coating disbondment thickness plays an essential role in CP shielding.

The CP shielding by disbonded coating is primarily attributed to the blocking effect of coating

disbondment on CP current. Under narrow disbonding gaps, the distribution of CP current is highly

non-uniform at the open holiday and under the disbonded coating. This effect is further enhanced

by limited diffusion of conductive ionic species through the thin solution layer trapped under the

coating. Thus, although the open holiday is under full CP, the disbonded region is shielded from

CP either partially or completely. With the increase in disbonding thickness, the distribution of CP

current can be improved around the holiday. The increased solution volume under the wider

coating disbondment enhances the diffusion of species, facilitating the CP permeation into the

disbondment. Thus, the CP shielding effect depends not only on the disbonding geometry, but also

on the conductivity of the trapped solution under coating.

The CP shielding by coating disbondment can result in cathodic polarization of steel at the

holiday, while the steel at the disbondment bottom can be at its corrosion potential, depending on

the disbonding thickness and applied CP. The potential difference produces separate anode and

cathode sites. The cathodic reaction occurs at the holiday, and the anodic reaction at the

disbondment bottom. The disbondment can become full of corrosion product, which is difficult to

diffuse through, further increasing the blocking effect on CP permeation. This is the key

114

mechanism resulting in localized corrosion on pipelines that are under CP. This phenomenon has

been demonstrated by frequent field experiences that extensive corrosion pits are found under

disbonded coating on a cathodically protected pipeline [143].

6.5 Summary

CP can be shielded by coating disbondment. With the increase in disbonding depth toward the

disbondment bottom, the shielding effect is more apparent. Besides, the potential monitoring test

demonstrated that the CP shielding can be mitigated by more negative CP potentials.

The geometrical factor of the coating disbondment plays an essential role in CP shielding.

When the disbondment becomes wider, the shielding effect is mitigated.

The CP shielding can result in separate anodic and cathodic reactions, which occur at the

disbondment bottom and the open holiday, respectively. This is the key mechanism that causes

localized corrosion under disbonded coating on a cathodically protected pipeline.

115

Chapter Seven: Cathodic Protection Shielding under Coating Disbondment on Pipelines in

the Absence of AC Interference5

7.1 Introduction

Pipelines are protected from external corrosion by both coating and cathodic protection (CP).

While the coating provides the first line of protective barrier, CP serves as a backup to prevent

corrosion attack either at coating faults such as pinholes and holidays or under coating

disbondment [144-146]. However, pipeline corrosion and corrosion induced cracking have been

widely documented [147-149]. When coating is disbonded at small faults, such as pinholes or

holidays, CP current can be shielded, either fully or partially, to reach the disbonding crevice,

especially at the crevice bottom. As a result, the CP could not protect the area that is in a corrosive

environment. This is called ‘CP shielding’. Another scenario encountered in reality to cause CP

shielding is the disbondment of a defect-free coating due to either an inadequate coating

application process or the lost adhesion of the coating to the steel substrate during service. For

example, spirally wrapped tape coatings can experience disbonding over pipeline welds. In this

situation, the CP shielding is attributed to the coating property [150]. It has been reported [151]

that up to 85% of all the external corrosion of pipelines are occurred under disbonded CP shielding

coatings. Non-shielding coatings are those which do not prevent distribution of CP current to the

steel substrate through the disbonded coating [152].

5 This work has been published in Corrosion Science (Da Kuang, Y. Frank Cheng, Study of cathodic protection

shielding under coating disbondment on pipelines, Corrosion Science 99, 2015, 249-257.)

116

It is generally accepted [153] that the widely used FBE and coal-tar enamel coatings exhibits

evidence of passage for CP current, while the high performance composite coating (HPCC) and

PE tape block CP current in long-term tests [153]. There have been works to measure and model

environmental chemistry and electrochemistry of the electrolyte trapped under CP shielding or

non-shielding coatings [154-157]. Moreover, lab tests have been conducted attempting to

understand the CP shielding behavior of some pipeline coatings through electrochemical

measurements, including potentiostatic current and electrochemical impedance spectroscopy [158-

160]. However, the majority of the testing did not duplicate the reality of pipeline corrosion under

disbonded coating in the field. For example, the solution pH trapped under the disbonded coating

was generally measured ex-situ by sampling small amount of electrolyte [161, 162].

In this work, a test rig was designed and constructed to simulate disbondment of defect-free

coatings on pipelines encountered in reality. The pH of the electrolyte under disbonded coating

was monitored in-situ by a micro pH meter which was installed into the disbondment crevice. Two

pipeline coatings, i.e., FBE and high-density polyethylene (HDPE), were used in this work.

Measurements of potential and potentiostatic current, optical characterization, and FT-IR were

combined to study the mechanistic aspects associated with the compatibility of coatings with

cathodic protection.

7.2 Experimental


The FBE and HDPE membranes, with a thickness of 250 µm and 850 µm, respectively, were

used in this work. All coating membranes, supplied by Bredero Shaw, were cut into circular shape

with a diameter of 5 cm.

117

The steel used in this work was X65 pipeline steel, with a chemical composition (wt.%): C

0.04, Si 0.20, Mn 1.5, P 0.011, S 0.003, Mo 0.020 and Fe balance. The steel coupons were sealed

with LECO epoxy, leaving a working area of 1 cm2. Prior to test, the electrode was subsequently

ground with 400, 800, 1000 and 1200 grit sand papers, and cleaned in distilled water and acetone.

A diluted bicarbonate solution, i.e., 0.01 M NaHCO3, was used to simulate the ground water

under the disbondment. It was made from analytic grade reagent and ultra-pure water, with a pH

of 7.5. All tests were conducted at 23C.

7.2.2 Measurements of CP permeability of coatings

A test rig was home-designed and constructed for testing of the CP permeability of coatings,

as shown in Fig. 7.1, where two test chambers were separated by a coating membrane. This setup

was to measure the compatibility of defect-free coatings with CP. A carbon rod was placed in the

top chamber and used as the auxiliary anode. The steel electrode was installed in the bottom

chamber. Two CP potentials of −0.85 V (copper/copper sulphate, CCS) and −1.0 V (CCS) were

applied, respectively, through an external DC power supply. A gap between the steel electrode and

the coating was created to simulate a disbondment crevice, which was 1.5 mm in height. A Loctite-

495 instant adhesive was used to seal the coating with the two cylindrical chambers.

Electrochemical measurements were conducted using a Solartron 1280C system, where a

platinum sheet and a saturated calomel electrode (SCE), installed in the top chamber in Fig. 7.1,

were used as counter electrode and reference electrode, respectively, and the steel electrode

installed in the bottom chamber was used as working electrode. Both chambers contained the

prepared 0.01 M NaHCO3 solution. The potential of the steel electrode and the potentiostatic

current between working electrode and counter electrode under CP application were measured.

118

Moreover, an Orion 9810BN micro pH electrode was installed in the disbondment between the

coating membrane and steel electrode to monitor the solution pH.

After testing, the morphology of the steel electrode was observed by an optical microscope.

Fig. 7.1. Schematic diagram of the experimental setup to measure the permeability of

coatings to CP, where WE and RE refer to working electrode and reference electrode,

respectively.

7.2.3 Coating characterization

The coating membranes were characterized by FT-IR with a Nicolet iS10 Spectrometer. For

all the spectra recorded, the samples underwent a 64-scan data accumulation in the range of 500 –

5500 cm−1 at a spectra resolution of 4.0 cm−1.

7.3 Results

7.3.1 Morphological observation of steel electrodes

Fig. 7.2 shows the optical morphology of the steel electrode in 0.01 M NaHCO3 solution

trapped under the disbonded HDPE coating after various times of testing at an applied CP potential

119

of −0.85 V (CCS). It is seen that the steel experiences serious corrosion after 5 days of testing.

With the increase of the testing time, more corrosion products are generated and deposited on the

steel surface. Obviously, the applied CP cannot provide effective protection to the steel under

disbonded HDPE coating.


under the disbonded HDPE coating after various times of testing at an applied CP

potential of -0.85 V (CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days.

120

Fig. 7.3 shows the optical morphology of the steel electrode in 0.01 M NaHCO3 solution

trapped under the disbonded FBE coating after various times of testing at an applied CP potential

of −0.85 V (CCS).


under the disbonded FBE coating after various times of testing at an applied CP potential

of -0.85 V(CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days.

121

It is seen that mild corrosion occurs on the steel electrode. Particularly, the corrosion product

generated is much less than that observed in Fig. 7.2, where the steel corrodes seriously under

disbonded HDPE coating. A few isolated corrosion pits are formed on the electrode surface. The

pits become bigger with the increase of the testing time. It is thus seen that FBE can permit CP

current penetrating to protect the steel from corrosion. However, under the given testing condition,

the steel under the disbonded FBE cannot be fully protected by the applied CP, as evidenced by

occurrence of localized corrosion.

To further investigate the effect of the elevated CP level on corrosion protection to steel under

disbonded FBE, a more negative CP potential of −1.0 V (CCS) is applied. The optical morphology

of the steel electrode after testing is shown in Fig. 7.4. It is seen that there is no corrosion occurring

on the electrode after 5 days of testing. There is no obvious corrosion sign observed within the

testing time period, even after 30 days. This indicates that cathodic protection current penetrates

through the FBE coating to fully protect the steel.

122


under the disbonded FBE coating after various times of testing at an applied CP potential

of -1.0 V (CCS): (a) 5 days; (b) 10 days; (c) 20 days and (d) 30 days.

123

7.3.2 Electrochemical measurements

Fig. 7.5 shows the potential of the steel in 0.01 M NaHCO3 solution trapped under the

disbonded HDPE and FBE coatings under CP potentials of −0.85 V (CCS) and −1.0 V (CCS),

respectively.

0 400 800 1200 1600 2000-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

30d

20d

5d

10d

1d

Time (s)

Po

ten

tia

l (V

,CC

S)

(a)

0 400 800 1200 1600 2000-0.40

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10

(b)

30d

20d

5d

10d

1d

Po

ten

tia

l (V

,CC

S)

Time (s)

124

0 400 800 1200 1600 2000-0.80

-0.75

-0.70

0.00

0.05

(c)

30d

1d

Time (s)

Po

ten

tia

l (V

,CC

S)

10d

5d

20d

Fig. 7.5. The potential of steel electrode in 0.01 M NaHCO3 solution trapped under the

disbonded HDPE and FBE coatings under CP potentials of -0.85 V (CCS) and -1.0 V

(CCS), respectively: (a) HDPE, -0.85 V (CCS); (b) FBE, -0.85 V (CCS); (c) FBE, -1.0 V

(CCS).

It is noted that the potential recorded includes the potential of the steel in the solution plus the

IR drop across the coating membrane while CP potentials are applied. It is seen in Fig. 7.5a that,

even under the CP potential of −0.85 V (CCS), the potential of the steel under the HDPE is positive

in the beginning. The potential is shifted negatively with time, and is about −0.065 V (CCS) only

after 30 days. Thus, the CP current could not permeate through the HDPE to cathodically polarize

the steel. For FBE coating, the potential of the steel is much more negative at individual time, as

shown in Fig. 7.5b. For example, the potential of the steel reaches −0.35 V (CCS) after 30 days,

compared to −0.065 V (CCS) for HDPE. Apparently, FBE is more compatible with CP in terms of

125

the permeation of cathodic protection current through the coating. It is also noted that, although

the CP current can permeate through the FBE, it does not fully penetrate the coating within the

time period of testing.

When the applied CP level is more negative up to −1.0 V (CCS), the potential of the steel is

shifted rapidly to the negative direction, as shown in Fig. 7.5c. The potential reaches to −0.70 V

(CCS) after 10 days and to −0.75 V (CCS) after 30 days of CP application. Thus, a more negative

CP level is helpful for cathodic protection current permeation through the coating.

Fig. 7.6 shows the potentiostatic current measured on steel in 0.01 M NaHCO3 solution trapped

under the disbonded HDPE and FBE coatings under CP potentials of −0.85 V (CCS) and −1.0 V

(CCS), respectively. It can be seen from Fig. 7.6a that anodic current densities are observed on

HDPE when the CP of −0.85 V (CCS) is applied. The HDPE coating generally behaves like an

ideal capacitor. The measured anodic current density is associated with the capacitive behavior of

the coating. It indicates that the applied CP cannot penetrate through HDPE. For FBE coating as

shown in Fig. 7.6b, the anodic current density obtained at the first day is due to the capacitive

behavior of the coating, and the CP is not yet to penetrate through the coating. With the increasing

time, the cathodic current density is recorded, indicating that CP goes through the coating to

cathodically polarize the steel. When a more negative CP potential of −1.0V (CCS) is applied, the

CP permeation is enhanced, as evidenced by the larger cathodic current density compared to that

measured at the CP of −0.85 V (CCS) at individual time, as shown in Fig. 7.6c. Moreover, it is

found that, even at the CP potential of −1.0 V (CCS), the anodic current density is obtained after

1 day of testing. Thus, even the FBE is CP permeable, it will take time for cathodic protection

current to penetrate.

126

0 400 800 1200 1600 20000.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20(a)

30d20d

5d

10d

1d

Time (s)

Cu

rre

nt

de

nsity (A

/cm

2)

0 400 800 1200 1600 2000-1.5

-1.2

-0.9

-0.6

-0.3

0.0

0.3

0.6

0.9(b)

30d

20d

5d10d

1d

Time (s)

Cu

rre

nt

de

nsity (A

/cm

2)

127

0 400 800 1200 1600 2000-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

(c)

5d

30d

20d

10d

1d

Time (s)

Cu

rre

nt

de

nsity (A

/cm

2)

Fig. 7.6. Potentiostatic current measured on steel in 0.01 M NaHCO3 solution trapped

under the disbonded HDPE and FBE coatings under CP potentials of -0.85 V (CCS) and -

1.0 V (CCS), respectively: (a) HDPE, -0.85 V (CCS); (b) FBE, -0.85 V (CCS); (c) FBE, -1.0

V (CCS).

7.3.3 In-situ pH monitoring under disbonded coating

Fig. 7.7 shows the time dependence of pH of the solution under HDPE and FBE coatings at

−0.85 V (CCS) and −1.0 V (CCS) of cathodic protection potentials, respectively. It is seen that the

solution pH under the HDPE disbondment fluctuates between 7.5 and 8 within the 30 days period

of testing, as shown in Fig. 7.7a. For FBE coating, the pH of the trapped solution increases

continuously with time, and is up to 9.3 after 30 days (Fig. 7.7b). When a more negative CP

potential of −1.0 V (CCS) is applied on the steel electrode under disbonded FBE coating, it can be

128

seen from Fig. 7.7c that the pH of the trapped solution increases more quickly, and reaches up to

10.4 after 30 days.

0 5 10 15 20 25 307.0

7.5

8.0

8.5

9.0

9.5

(a)

Time (day)

pH

0 5 10 15 20 25 307.0

7.5

8.0

8.5

9.0

9.5

(b)

pH

Time (day)

129

0 5 10 15 20 25 307.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

(c)

Time (day)

pH

Fig. 7.7. Time dependence of pH of the solution under HDPE and FBE coatings at -0.85

V(CCS) and -1.00 V(CCS) CP potentials, respectively: (a) HDPE, -0.85 V(CCS); (b) FBE, -

0.85 V(CCS); (c) FBE, -1.0 V(CCS).

7.3.4 FT-IR characterization of the coatings

Fig. 7.8 shows the FT-IR spectrum of the HDPE coating before and after 30 days of testing

under CP of −0.85 V (CCS). The characteristic functional groups identified are listed in Table 7.1.

It is seen that there is no difference of the HDPE spectra obtained before and after the testing.

Moreover, the positions of the characteristic functional groups, including C-C, C-H2 and C-H3, are

consistent with others’ results [163-165].

130

4000 3000 2000 1000 0

1080 cm-1

2850 cm-1

720 cm-1

1470 cm-1

wavenumber (cm-1)

ab

so

rba

nce

HDPE

HDPE after 30 days

2965 cm-1

Fig. 7.8. FT-IR spectrum of the HDPE coating before and after 30 days of testing under

CP of -0.85 V (CCS).

Table 7.1. Characteristic functional groups of HDPE identified in the FT-IR spectrum in

Fig. 7.8.

Band (cm-1) Groups

2965

2850

1470

1080

720

C-H3 stretching [163]

C-H2 stretching [164]

C-H2 bending [165]

C-C stretching [163]

C-H2 scissoring [165]

131

Fig. 7.9 shows the FT-IR spectrum of the FBE coating before and after 30 days of testing under

CP of −0.85 V (CCS). The characteristic functional groups identified are listed in Table 7.2. It is

seen that three characteristic absorption peaks of the oxirane ring are observed. The peak at 831

cm-1 is attributed to the C-O band of the oxirane group. The peak located at 2873 cm-1 is due to the

C-H tension of methylene group in the epoxy ring. The next peak is located at 4623 cm−1, which

is a combination band of the second overtone of epoxy ring stretching with the fundamental C-H

stretching at 2873 cm−1 [128]. The results are well consistent with others’ [166, 167].

6000 5000 4000 3000 2000 1000 0

831 cm-1

1608 cm-1

2873 cm-1

4066 cm-1

4623 cm-1

1509 cm-1

3600 cm-1

5215 cm-1

Ab

so

rba

nce

Wavenumber (cm-1)

FBE 0day

FBE 5days

FBE 10days

FBE 20days

FBE 30days

Fig. 7.9. FT-IR spectrum of the FBE coating before and after 30 days of testing under CP

of -0.85 V (CCS).

132

Table 7.2. Characteristic functional groups of FBE identified in the FT-IR spectrum in Fig.

7.9.

Band (cm-1) Groups

5215

4623

4066

3600

2873

1608

1509

831

Combination asymmetric stretching and bending of O-H [129, 130]

Overtone of C-H stretching of the aromatic ring [131]

Stretching C-H of aromatic ring [132]

O-H stretching [133]

C-H stretching [128]

Stretching C=C of aromatic rings [166]

Stretching C-C of aromatic rings [167]

Stretching C-O-C of oxirane group [167]

Moreover, two types of water bands are identified in the spectra, i.e., highly mobile free water

molecules at about 3600 cm-1, and the combination of asymmetric stretching and bending of

hydroxyl vibrations at 5215 cm-1 [168]. There are obvious changes of the absorbance intensities

of functional groups along with time. The absorption band of C-H stretching located at 2873 cm-1

becomes stronger and the band of C-H overtone stretching of the aromatic ring at 4623 cm-1

becomes weaker with the increasing time. Another important change is the intensity of the bands

of hydroxyl groups. The peaks of 3600 cm-1 and 5215 cm-1, representing the absorbed water in

FBE coating, become sharper with the increasing time, indicating that water uptakes occur

continuously with time.

133

7.4 Discussion

7.4.1 Shielding effect of HDPE coating on CP

The HDPE is a thermoplastic hydrocarbon material with a high molecular weight. It is a widely

used pipeline coating due to its large electrical resistance and low water absorption in practice. In

the present work, the potential and potentiostatic current measurements show that the applied

cathodic protection current cannot penetrate into the HDPE coating membrane to reach the steel

surface. This conclusion is further demonstrated by the in-situ pH measurements of the trapped

solution under disbonded HDPE. It is generally accepted [138] that the electrochemical reduction

of water is enhanced by CP, resulting in generation of hydroxyl ions to elevate the solution pH.

There is little change of the pH of the trapped solution under HDPE coating. Moreover, extensive

corrosion occurs on the steel electrode in the solution under the HDPE, although the cathodic

protection potential of -0.85 V (CCS) is applied. Thus, there is no cathodic protection current

reaching the steel through HDPE for corrosion protection. The corrosion product deposits due to

limited diffusion under the coating disbondment. All results obtained support the conclusion that

the HDPE is a CP shielding coating, which blocks cathodic protection current to reach the steel

for corrosion protection if the coating is disbonded and a corrosive environment is developed under

the coating.

The FT-IR results about the molecular structure of HDPE show that the chemical structure of

HDPE does not change before and after testing. Moreover, there is no characteristic peak specific

to hydroxyl groups identified. This shows that water molecules are not absorbed into the HDPE.

It has been acknowledged [169] that water permeation into a hydrocarbon coating is predominated

by its polarity. The HDPE is a non-polar polymer, and is highly resistant to permeation of water.

It has been tested that the HDPE can maintain a high resistivity after a long-term wetting, and the

134

coating impedance is up to 1017 (Ω∙cm) [170]. Thus, it is the non-polar structure of HDPE that

makes water molecule impermeable to the coating. The high resistivity of the coating blocks the

cathodic protection current circuit. As a result, the HDPE coating shields CP from reaching the

steel for corrosion protection.

7.4.2 Permeability of FBE coating on CP

The present work demonstrates that FBE is permeable to CP, as confirmed by the

measurements of the potential and potentiostatic current of the steel electrode in the test solution

trapped under the disbonded coatings, as well as monitoring of solution pH. The morphology of

the steel electrode also confirms that CP can be permeable to FBE to reach the steel for corrosion

protection. At the same time, it is found in this work that, within the testing period, CP is partially

permeable through the FBE coating. In other words, the applied CP potential, i.e., -0.85 V (CCS),

cannot provide a full protection to steel in corrosive environments trapped under disbonded FBE

coating. Thus, localized corrosion can occur on the steel. This finding is interesting since FBE has

been regarded as a coating with ‘fail safe’ characteristics in the presence of cathodic protection.

Due to its CP compatibility, the steel can be protected even it is in a corrosive environment under

disbonded coating. This work demonstrates that the steel is actually under a partial cathodic

protection under disbonded FBE. Moreover, this work finds that the CP permeation through FBE

is time dependent. As time is elapsed, there is an increased CP permeating through the FBE to

reach the steel.

A conceptual model is developed to illustrate the mechanistic aspect of the cathodic protection

compatibility of FBE coating, as shown in Fig. 7.10. FBE is a highly polar coating containing

135

hydroxyl groups. Water is also a polar molecule and capable of forming hydrogen bonds with the

hydroxyl groups.

Fig. 7.10. Schematic diagram of the model for illustrating the compatibility of FBE coating

with CP.

Three steps are included in the proposed model. In step one, a monolayer of water is formed

on the coating surface by bonding with the polar species in the coating. In step 2, the absorbed

water molecules diffuse through the micropores and voids existing in the cross-linked structure of

the FBE, as shown in Fig. 5.11. As accumulating at the micropores, the water molecules further

diffuse into the polymeric network and enter into the densely cross-linked parts through

connections of adjacent micropores. The interactions of water along with these pores cause

136

disruption of hydrogen bonds, resulting in creation of more micro-voids or pores accessible to

water. Thus, with the increase of time, water could diffuse into the polymeric matrix via more

micropores, which serve as effective ionic channels for the cathodic protection current. Thus, in

step 3, under the cathodic protection application, water and ions would permeate through the

coating via the micropore channels driven by the effect of electroendosmosis. As shown in the

model, ions with opposite charge to the steel could pass through the channels, and reach the steel

surface. Along with the time, more micropores are filled with ions, resulting in a dramatically

decrease of the resistivity of the coating. It was reported [170] that the resistance of FBE drops

tremendously from 1015 to 109 Ω∙cm after 500 h of water absorption testing, resulting in the FBE

coating changing from nonconductive to semi-conductive or conductive, along with the CP

permeation.

The rate of water transfer through a polymeric coating is proportional to the applied cathodic

protection potential [171]. The more negative the cathodic protection potential, the more cations

diffuse through the coating to decrease its resistivity. Thus, more cathodic protection current

permeates through the coating for corrosion protection. Moreover, it was proposed [172] that the

ion-transport process inside a coating is not homogeneous, with some regions where the ionic

transport occurs with a high resistance, and others where ions transports with a relatively low

resistance. In this model, the cathodic protection current permeates preferentially through the low-

resistance regions to reach the steel, causing an uneven distribution of the protective current on the

steel surface. Thus, localized corrosion occurs at areas with an insufficient cathodic protection.

However, at sufficiently negative cathodic protection potentials, such as -1.0 V(CCS), the

enhanced electroendosmosis is able to drive cations to diffuse through the coating at both low-

137

resistance and high-resistance regions in the coating, generating a uniform distribution of the

cathodic current on the steel for corrosion protection under the disbonded coating

It should be noted that the focus of this work is to evaluate the CP permeation through pipeline

coatings, such as FBE and HDPE, for corrosion protection when a corrosive electrolyte is

generated under disbonded coating. This work does not consider the CP current consumption on

different coatings. From the interactions with pipeline industry, the pipeline integrity engineers

care if external corrosion of pipelines can be controlled effectively by both coating and CP. It

seems that the CP current consumption is not in the list of top priority in integrity maintenance.

However, the CP current consumption on pipelines coated with different coatings is also an

important question on managing the pipeline safety effectively and efficiently. Furthermore,

previous work conducted in Chapter Six has demonstrated that CP could enhance coating

disbondment at defects such as pinholes or holidays. For defect-free coatings, there has been

limited work to study how the CP passing affects the coating adhesion. The present work focuses

on CP permeation through an already-disbonded coating, where the disbonding is not driven by

CP, but by coating application or other factors.

7.5 Summary

The HDPE is a CP-shielding coating, which blocks cathodic protection current to reach the

steel for corrosion protection if the coating is disbonded and a corrosive environment is developed

under the coating. The FBE is permeable to CP. Thus, CP could penetrate the coating to protect

the steel from corrosion under disbonded coating.

Online monitoring of pH of the electrolyte trapped under disbonded coating provides a

convenient, promising method to indicate if CP shielding occurs under coating disbondment. The

138

pH results are well consistent with the measurements of potential and potentiostatic current

density.

As confirmed by the FT-IR characterization, after CP permeating testing, the chemical

structure of HDPE does not change, and water molecules are not absorbed into the coating.

Conversely, there are obvious changes of the absorbance intensities of functional groups of FBE

under cathodic protection application in the testing. Thus, the CP compatibility of pipeline coatings

can be evaluated from the perspective of molecular structure.

The CP permeation through the FBE coating is time dependent. Over the testing period in this

work, i.e., after 30 days of testing, the CP is partially permeable through the coating. In other

words, the steel under the disbonded FBE coating is not under full protection. A more negative CP

potential level, i.e., -1.0 V (CCS), enhances the CP permeation.

The CP permeation through the FBE coating occurs via three stages, i.e., bonding of water

molecules with the polar species in the coating, accumulation of water molecules at accessible

micropores existing in the coating, and permeation of water and ions through the coating via the

micropore channels driven by electroendosmosis introduced by the cathodic protection.

139

Chapter Eight: Cathodic Protection Shielding under Coating Disbondment on Pipelines in

the Presence of AC Interference6

8.1 Introduction

As discussed previously in Chapter Seven, CP can become shielded from reaching the bottom

of coating disbondment. As a result, corrosion occurs under the disbonded coating once an

electrolyte is trapped underneath, while the pipeline is under CP. At the same time, the presence

of AC interference from the adjacent high-voltage power transmission lines is expected to affect

the CP permeation into the coating disbondment as the AC could enhance the disbonding of

coating from pipelines and accelerate corrosion of the steel [173]. To date, there has been limited

work to investigate the CP permeation into the disbonding crevice under the coating in the presence

of AC interference.

In this work, electrochemical measurements and pH monitoring were conducted in the

simulated electrolyte trapped under disbonded coating to study the role of AC in CP permeation

into the disbonding crevice. The parametric effects, such as AC current density, CP potential and

disbonding thickness, on CP permeation were determined. The mechanistic aspect regarding the

effect of AC on CP shielding was discussed.

6 This work has been published in Corrosion Engineering, Science and Technology (D. Kuang, Y.F. Cheng, Effect of

alternating current interference on coating disbondment and cathodic protection shielding on pipelines, Corrosion

Engineering Science and Technology 50, 2015, 211-217).

140

8.2 Experimental


FBE coated X65 steel coupons, which were supplied by Bredero Shaw, were used in this work.

The thickness of the coating was 180 μm. The chemical composition (wt.%) of the steel is C 0.04,

Si 0.2, Mn 1.5, P 0.011, S 0.003, Mo 0.02 and Fe balance.

For testing of CP permeation, a neutral pH bicarbonate solution simulating the trapped

electrolyte under the disbonded coating was used. The solution contained 0.01 M NaHCO3 and

was purged with 5% CO2/N2 for 48 h prior to and throughout testing. The solution pH was 7.5. All

solutions were made from analytic grade reagent and ultra-pure water.

8.2.2 CP permeation measurements in the presence of AC interference

The setup shown in Fig. 6.1 was used to simulate the crevice under a disbonded coating and

the installation of microprobes for in-situ potential and pH measurements. The X65 steel substrate

had a dimension of 200 mm × 25 mm × 20 mm. To prepare an artificial disbondment, a FBE

membrane with a thickness of 200 μm was applied on the steel surface using a TESA double-sided

self-adhesive tape. A hole (10 mm in diameter) was created on the coating to simulate the coating

holiday. Six smaller holes with diameter of 3.16 mm were created in order to install the

microprobes, which were 30 mm, 60 mm, 90 mm, 120 mm, 150 mm and 180 mm from the bigger

hole, respectively. The CP potential and AC current density were applied at the bigger holiday.

At each smaller hole, a combination of measurement kit including a saturated calomel electrode

(SCE) reference electrode and a pH meter was installed to monitor local potential and solution pH.

The potential measurement was performed through the AC/DC real-time data acquisition system.

141

The local solution pH was measured using an Oaklon Acorn pH meter. The testing was performed

at room temperature of about 23 oC.

To prepare the simulated disbondment, two layers of tape with a known thickness were adhered

to the surface of the steel. The FBE membrane was adhered to form an artificial disbondment on

the steel surface. The disbonding thickness was determined by the total thickness of the tape layers,

and was measured by a coating thickness gauge. The boundaries of tape/steel/coating were sealed

with epoxy resin.

8.3 Results

8.3.1 Potential distribution under the coating disbondment

Fig. 8.1 shows the potential distribution under the coating disbondment (the disbonding

thickness is 120 μm) at CP potential of -0.875 V (SCE) under various AC current densities. It is

seen that, even in the absence of AC, CP would be shielded to reach the disbonding crevice along

the depth direction. At the open holiday, i.e., the distance to the holiday of 0 mm, the measured

potential is at the applied CP potential of -0.875 V (SCE) up to 48 h of testing. However, at the

distance of 30 mm from the open holiday, the potential is shifted less negatively to about -0.80 V

(SCE). With the increase in distance from the holiday, the potential becomes less negative. When

the distance is up to 60 mm, the potential would not change with the increase of the distance to the

holiday. It is thus seen that the geometry and coating disbondment shields the CP permeation,

decreasing the CP performance at the crevice depth. The critical depth in the absence of AC is 60

mm, above which CP would not be able to reach, and the steel is at free corrosion potential of -

0.755 V (SCE).

142

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(a)

Time (h)

P

ote

ntial (V

, S

CE

)

0 mm

30 mm

60 mm

90 mm

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(b)

Time (h)

Pote

ntial (V

, S

CE

)

0 mm

30 mm

60 mm

90 mm

143

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(c) 0 mm

30 mm

60 mm

90 mm

Time (h)

Pote

ntial (V

, S

CE

)

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(d) 0 mm

30 mm

60 mm

90 mm

Time (h)

Pote

ntial (V

, S

CE

)

144

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(e) 0 mm

30 mm

60 mm

90 mm

Time (h)

Pote

ntial (V

, S

CE

)

0 10 20 30 40 50-1.1

-1.0

-0.9

-0.8

-0.7

-0.6

-0.5

120 mm

150 mm

180 mm

(f) 0 mm

30 mm

60 mm

90 mm

Time (h)

Pote

ntial (V

, S

CE

)

Fig. 8.1. Local potential profile under the coating disbondment (disbonding thickness of

120 μm) at CP potential of -0.875 V (SCE) under various AC current densities (a) 0 A/m2,

(b) 100 A/m2, (c) 200 A/m2, (d) 300 A/m2, (e) 400 Acm2, (f) 500 A/m2.

145

Upon AC application, the DC potential of the steel electrode is shifted negatively. Moreover,

as the AC current density increases, the potential shift is more negative, as seen from the potential

measurements on the holiday (distance of 0 mm) in Figs. 8.1b-f. Moreover, along the direction

towards the disbonding depth, the potential is shifted less negatively. Obviously, the CP shielding

due to the crevice geometry also applies under the AC interference. It is also interesting to find

that, as small AC current densities, such as 100 A/m2, the critical crevice depth where CP cannot

reach is about 150 mm from the open holiday. With the increasing AC current density, the critical

crevice depth decreases. When AC current density is up to 200 A/m2, 300 A/m2 and 400 A/m2, the

critical depths are 120 mm, 90 mm and 30 mm, respectively. When the AC current density is 500

A/m2, there is no CP that can permeate into the potential-monitoring regions. Thus, at small AC

current densities, the AC results in more negative potentials at individual disbonding depth

compared to those measured in the absence of AC. However, the trend is diminished with

increasing AC current density. When compare the potential profiles measured in the absence of

AC (Fig. 8.1a) and at AC current density of 400 A/m2, there is a quite high similarity under the

disbonding crevice, except the latter has a more negative potential at open holiday. At large AC

current density such as 500 A/m2, CP is completely shielded from reaching the disbonding crevice.

8.3.2 Solution pH distribution under the coating disbondment

Fig. 8.2 shows the solution pH profile under the coating disbondment (disbonding thickness of

120 μm) at CP potential of -0.875 V (SCE) under various AC current densities. It is seen that, in

the absence of AC, the CP application enhances the solution pH at the open holiday to about 8.4

(the original solution pH is 7.5).

146

0 10 20 30 40 506

7

8

9

10

11

12

120 mm

150 mm

180 mm

(a)

Time (h)

0 mm

30 mm

60 mm

90 mm

pH

0 10 20 30 40 506

7

8

9

10

11

12

120 mm

150 mm

180 mm

(b)

Time (h)

0 mm

30 mm

60 mm

90 mm

pH

147

0 10 20 30 40 506

7

8

9

10

11

12

120 mm

150 mm

180 mm

Time (h)

(c) 0 mm

30 mm

60 mm

90 mm

pH

0 10 20 30 40 506

7

8

9

10

11

12

120 mm

150 mm

180 mm

(d)

Time (h)

0 mm

30 mm

60 mm

90 mm

pH

148

0 10 20 30 40 506

7

8

9

10

11

12

120 mm

150 mm

180 mm

Time (h)

pH

(e) 0 mm

30 mm

60 mm

90 mm

0 10 20 30 40 506

7

8

9

10

11

12

13

120 mm

150 mm

180 mm

(f)

Time (h)

pH

0 mm

30 mm

60 mm

90 mm

Fig. 8.2. Local pH profile under the coating disbondment (disbonding thickness of 120 μm)

at CP potential of -0.875 V (SCE) under various AC current densities (a) 0 A/m2, (b) 100

A/m2, (c) 200 A/m2, (d) 300 A/m2, (e) 400 A/m2, (f) 500 A/m2.

149

Under the disbonding crevice, there is a lower solution pH compared to that at the holiday.

With the increase in disbonding depth, the pH is decreased. Upon AC application, the solution pH

at open holiday is elevated obviously, but the solution pH inside the disbonding crevice is lower.

As the distance to the open holiday increases towards the disbondment bottom, the pH further

drops. Moreover, with the increase in AC current density, the solution pH under the disbonding

crevice decreases.

8.4 Discussion

8.4.1 Effect of AC current density on coating disbondment

Previous work in Chapter Five presents that AC could enhance disbonding of coating from the

steel substrate through the coating disbondment (CD) tests, as shown in Figs. 5.3 to 5.6. To

understand mechanistically this phenomenon, the effect of AC current density on local pH inside

the disbonding crevice, e.g., at 180 mm from the open holiday, is plotted in Fig. 8.3. It is clear that

the solution pH at the crevice bottom is elevated by increasing AC current density.

150

0 10 20 30 40 506.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

Time (h)

0 A/m2

100 A/m2

200 A/m2

300 A/m2

400 A/m2

500 A/m2

pH

Fig. 8.3. Effect of AC current density on local pH inside the disbonding crevice at 180 mm

from the open holiday (disbondment thickness: 120 μm).

Upon application of AC, anodic oxidation and cathodic reduction reactions occur during the

positive and negative half-cycles, respectively. During corrosion of X65 pipeline steel in neutral

pH bicarbonate solution under disbonded coating, the anodic and cathodic reactions at the AC

positive and negative half cycles are the iron oxidation and reduction of water (in the absence of

oxygen in the environment), respectively [174]:

Fe → Fe2+ + 2e (8.1)

2H2O + 2e → H2 + 2OH- (8.2)

151

The electrochemical reactions generate OH- ions as product to elevate pH of solution trapped under

disbonded coating. Moreover, as the AC current density increases, the water reduction is enhanced

at more negative potential in AC negative cycles. Thus, the solution becomes more alkaline, as

shown in Figs. 8.2 and 8.3. Obviously, solution alkalization is the primary reason to result in

coating disbondment during AC application. The generated hydrogen bubbles due to cathodic

reaction (8.2) can also contribute to coating disbonding by weakening the bonding of the coating

to steel.

8.4.2 Effect of AC current density on CP permeation into disbonding crevice

Furthermore, it is demonstrated from this work that the potential of steel at the open holiday is

shifted negatively upon AC application, as seen in Fig. 8.1. This result is consistent with previous

studies in Chapters Three and Four. This is attributed to accelerated diffusion of cation ions that

are generated during corrosion of steel under AC enhanced electric field. As a consequence, the

number of positive charges in the double-charge layer decreases, resulting in a negative shift of

the potential. It is noted that AC induced negative shift of the potential is due to enhanced

corrosion, rather than cathodic polarization of the steel.

In the absence and presence of AC, CP can be shielded by the geometry of coating

disbondment, as shown in Fig. 8.1, where the local potentials inside the disbonding crevice are

less negative than the potential at the open holiday. In order to show clearly the effect of AC

current density on CP permeation into the disbonding crevice, the local potentials at the disbonding

sites of 30 mm and 180 mm from the holiday are plotted as a function of AC current density, as

shown in Fig. 8.4. It is seen that, at small AC current densities such as 100 A/m2 and 200 A/m2,

152

the local potential at both sites is shifted negatively by AC. As discussed, this shift is associated

with the enhanced corrosion of steel under AC interference. When the AC current densities

increase to 300 A/m2 and 500 A/m2, the potentials are shifted towards the positive direction. At

high AC current densities, the iron dissolution and water reduction are further enhanced in the

positive and negative cycles of AC. There are more ferrous and hydroxyl ions generated during

anodic and cathodic reactions. Due to limited space under the disbonding crevice, the corrosion

product are not able to diffuse outwards freely. The positive shift of the steel potential is associated

with deposit of corrosion product, such as Fe(OH)2, on the steel.

Furthermore, at small AC current densities, the AC enhanced steel corrosion generates ferrous

and hydroxyl ions, improving the conductivity of solution trapped under disbonded coating. This

results in enhancement of permeation of CP current into the crevice. As shown in Fig. 8.1, at AC

current densities of 100, 200 and 300 A/m2, the critical disbonding depths are 120 mm, 90 mm and

60 mm, respectively, while that in the absence of AC is 30 mm only. The reason that the critical

disbonding depth decreases with increasing AC current density is associated with the blocking

effect of corrosion product under coating. As discussed, corrosion product generates and deposits

at high AC current densities, blocking the ionic diffusion and CP current and decreasing the CP

permeation. When the AC current density is increased to 500 A/m2, the CP is blocked completely

to reach the disbonding crevice. Thus, it is important to note that, under high AC current densities,

the applied CP can be shielded completely from reaching the coating disbondment.

153

0 10 20 30 40 50-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50

Time (h)

Pote

ntial (V

, S

CE

)

(a) 0 A/m2

100 A/m2

200 A/m2

300 A/m2

400 A/m2

500 A/m2

0 10 20 30 40 50-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50

Time (h)

Pote

ntial (V

, S

CE

)

(b) 0 A/m2

100 A/m2

200 A/m2

300 A/m2

400 A/m2

500 A/m2

Fig. 8.4. Local potentials at the disbonding sites of (a) 30 mm and (b) 180 mm from the

holiday are plotted as a function of AC current density (CP potential: -0.875 V (SCE)

disbonding thickness: 120 μm).

154

8.5 Summary

The AC is able to enhance coating disbonding from the steel substrate, which is attributed to

the solution alkalization by the enhanced cathodic reduction reaction, resulting in generation of

hydroxyl ions.

In the absence and presence of AC, CP can be shielded by the geometry of coating

disbondment. At small AC current densities, such as 100 A/m2, the AC enhanced corrosion

generates ferrous and hydroxyl ions, improving the conductivity of solution trapped under

disbonded coating. This results in enhancement of permeation of CP current into the crevice.

However, with the increase of AC current density, corrosion product generates and deposits in the

solution, blocking the ionic diffusion and CP permeation. When the AC current density is up to

500 A/m2, the CP is blocked completely to reach the disbonding crevice. Thus, it is important to

note that, under high AC current densities, the applied CP can be shielded completely from

reaching the coating disbondment.

155

Chapter Nine: Effects of AC Interference on Cathodic Protection Potential and its

Effectiveness for Corrosion Protection on Pipelines

9.1 Introduction

The AC interference can affect, actually decrease, the effectiveness of CP on pipelines by

shifting the applied CP potential from the design value [86]. It was recognized [87] that the

recommended CP criteria by NACE International is not sufficient to provide full protection for

pipelines from corrosion in the presence of AC, especially at large AC current densities. The field

experiences also demonstrated that the AC induced corrosion can occur on carbon steels even

under higher CP levels [13-15]. As a result, new CP criteria according to field survey and

laboratory tests were developed for eliminating the risks of AC corrosion of pipelines [88].

However, the criteria did not apply at AC current density in excess of 70 A/m2. Fu and Cheng [87]

demonstrated that, if the AC current density is less than 20 A/m2, a CP potential of -0.95 V (CCS)

could provide a full protection over the steel. However, when the AC current density is higher than

20 A/m2, the CP potential should be shifted negatively to avoid corrosion of pipelines. It is thus

seen that a consensus in terms of the effect of AC on CP has been lacking.

The primary reason to cause the inability of quantifying the AC effect on CP performance is

due to the fact that the potential of the pipeline deviates from the applied CP level in the presence

of AC. It was reported [91] that CP monitoring in the presence of simultaneous AC interference

could lead to erroneous measurements. Experimental results [175] found that the DC potential of

pipeline steels under CP application depends on both the AC current density and the CP level. The

156

AC is able to make the DC potential of the steel more or less than the applied CP potential, but the

exact mechanism has remained unknown.

In this work, it is attempted to determine the ‘true’ DC potential of the pipeline steel which is

under simultaneous application of CP and AC. A home-developed voltage acquisition and

processing software was used to obtain the DC potential of a X65 pipeline steel from the recorded

mixed potential signals when various AC current densities and CP potential were applied on the

steel electrode in a near-neutral pH bicarbonate solution. The mechanistic aspect of the effect of

AC on shift of applied CP potentials was investigated by considering the double-charge layer at

the steel/solution interface under AC and CP. Moreover, weight-loss tests were conducted to

determine the corrosion rate of the steel, providing direct results on the effect of AC on CP

performance.

9.2 Experimental


Specimens used in this chapter were cut from a sheet of X65 pipeline steel, with a chemical

composition (wt.%): C 0.04, Si 0.2, Mn 1.5, P 0.011, S 0.003, Mo 0.02 and Fe balance. The steel

specimens were sealed with epoxy, leaving a working area of 1.0 cm2. The specimen preparation

was controlled carefully to avoid any groove and bubble generating at the epoxy/steel interface.

The prepare steel specimen was subsequently ground with 400, 800, 1000 and 1200 grit sand

papers, and cleaned in distilled water and acetone.

The test solution was a near-neutral pH bicarbonate solution, which has been widely used to

simulate the diluted electrolyte trapped disbonded coating [176]. The solution contained 0.483 g/L

NaHCO3 and it was made from analytic grade reagents and ultra-pure water. Prior to test, the

157

solution was purged with 5% CO2 balanced with N2 gas for 1 h to achieve an anaerobic and near-

neutral pH condition (pH = 7.5). The gas flow was maintained throughout the test. All tests were

conducted at 23.

9.2.2 DC potential measurements

The DC potential of the steel electrode under various AC current densities and CP potentials

were measured through the home-developed DAQ system, which is same as that used in previous

chapters, and the schematic diagram of the experimental setup is shown in Fig. 9.1, where the steel

specimen, which was used as WE, was connected to the negative pole of a DC power supply. A

platinum sheet used as CE was connected to the positive pole. A carbon rod was connected to an

AC signal source through a slide rheostat. The AC current density flowing between WE and the

carbon rod was varied from 0 to 200 A/m2 by adjusting the AC voltage applied on the rheostat. A

capacitor was used to block the CP current from flowing into the AC circuit, and an inductor was

installed between WE and the DC source to prevent the interference of AC on DC supply. A SCE

was used as RE, and the distance between RE and WE was about 1 mm in order to reduce the

ohmic drop. All potentials in this chapter were converted to CCS to be consistent with the industry

use. During the application of CP potentials on WE galvanostatically, the DAQ system measured

the DC potential of the steel WE relative to RE. For CP applied on WE potentiostatically, the

current was recorded by a Solartron 1280C electrochemical workstation.

158

Fig. 9.1. Schematic diagram of the experimental setup to measure the DC potential of the

steel electrode under various AC current densities and CP potentials.

9.2.3 Weight-loss tests

Before testing, steel coupons were weighed by an electronic balance with an accuracy of 0.1

mg. It was immersed in the test solution, where under various CP potentials, i.e., no CP, -0.850 V

(CCS), -0.925 V (CCS) and -1.0 V (CCS), and AC current densities, i.e., 0, 10, 50, 100 and 200

A/m2 were applied for 10 days. After testing, the corrosion products formed on the steel surface

were removed carefully by both mechanical and chemical methods according to ASTM G1-03

[177]. Light scraping and scrubbing were used to remove tightly adherent corrosion products, and

a descaling solution containing 500 mL HCl, 3.5 g hexamethylenetetramine and 500 mL distilled

water was used in chemical removal procedure. The cleaning process was repeated several times

to remove corrosion products thoroughly. After rinsing with distilled water and alcohol, the steel

159

coupons were dried and weighed again. The weight changes under specific testing conditions were

recorded.

Each test was repeated three times to obtain an average value under individual testing

condition. The corrosion rate (mm/year) of the steel was calculated from the weight-loss

measurements and the testing time.

9.3 Results

9.3.1 Measurements of DC potential of the steel

Figs. 9.2 to 9.4 show the DC potentials of the steel electrode at various AC current densities

under CP potentials of -0.850 V (CCS), -0.925 V (CCS) and -1.0 V (CCS), respectively, in NS4

solution. Four potential regions are included in all plots, i.e., (a) CP potential region: from 0 to 600

s, CP is applied on the steel; (b) CP + AC region: from 600 s to 1200 s, various AC current densities

are applied on the steel while the CP is maintained; (c) AC region: from 1200 s to 1800 s, the CP

is stopped, and AC is applied; and (d) open-circuit potential (OCP) region: after 1800 s, both CP

and AC are stopped. In the first region up to 600 s, the DC potential of the steel is at the applied

CP potential.

When the applied CP potential is -0.850 V (CCS), the DC potential is shifted negatively upon

addition of AC in the CP + AC region. Moreover, the DC potential becomes more negative with

the increasing AC current density. The most negative value is observed at AC current density of

200 A/m2. In the AC region where the CP is stopped, the DC potential is shifted positively. The

steady-state value of the DC potential is more negative with the increasing AC current density. In

the OCP region, when both CP and AC are stopped, the DC potentials tend to reach a steady value

of -0.785 V (CCS), i.e., the corrosion potential of X65 steel in NS4 solution.

160

0 500 1000 1500 2000 2500-1.05

-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

OCP

AC

CP+ACCP

Time (s)

Po

ten

tia

l (V

, C

CS

)

10 A/m2

50 A/m2

100 A/m2

200 A/m2


potential of -0.850 V (CCS) in NS4 solution

When the CP potential -0.925 V (CCS), there are identical features of the DC potential in AC

and OCP regions to those observed in Fig. 9.2, where the CP potential is -0.850 V (CCS). However,

in the CP + AC region, the DC potential becomes more positive than the CP potential at small AC

current densities of 10 and 50 A/m2. With the increase of AC current density to 100 A/m2, the DC

potential is shifted negatively, which is as same as that observed in Fig. 9.2.

161

0 500 1000 1500 2000 2500-1.05

-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

Time (s)

Po

ten

tia

l (V

, C

CS

)OCP

AC

CP+ACCP

10 A/m2

50 A/m2

100 A/m2

200 A/m2



Under the CP potential of -1.0 V (CCS), it is interesting to see that the DC potential is shifted

in the positive direction, which is opposite to those observed at CP potentials of -0.850 V (CCS)

and -0.925 V (CCS) in Figs. 9.2 and 9.3. Moreover, as the AC current density increases, the DC

potential becomes more negative. In the AC region, the DC potential is further shifted positively,

and reaches a steady-state value, which is more negative with the increase of AC current density.

When both CP and AC are stopped, the DC potential tends towards the OCP value.

162

0 500 1000 1500 2000 2500-1.05

-1.00

-0.95

-0.90

-0.85

-0.80

-0.75

-0.70

OCP

AC

CP+AC

CP

Time (s)

Po

ten

tia

l (V

, C

CS

)

10 A/m2

50 A/m2

100 A/m2

200 A/m2



9.3.2 Measurements of CP current density

Figs. 9.5 to 9.7 show the CP current densities of the steel electrode at various AC current

densities under CP potentials of -0.850 V (CCS), -0.925 V (CCS) and -1.0 V (CCS), respectively,

in NS4 solution. In these measurements, the CP is applied potentiostatically. Three regions are

observed in all plots, i.e., (a) CP-I region: from 0 to 800 s, the recorded current density is generated

under applied CP potential; (b) CP + AC region: from 800 s to 1600 s, various AC current densities

are applied while maintaining the CP on the electrode; and (c) CP-II region: from 1600 s to 2400

s, CP is continuously applied but AC is stopped. In the first region up to 800 s, cathodic current

densities of -0.2 A/m2, -0.4 A/m2 and -1.0 A/m2 are measured at individual CP potentials of -0.850

V (CCS), -0.925 V (CCS) and -1.0 V (CCS), respectively. The steel electrode is under full

163

protection under these CP levels. Moreover, in the CP-II region of all plots, the current densities

are back to the original values identical to those in the CP-I region.

When the applied CP potential is -0.850 V (CCS), in the CP + AC region, anodic current

densities are recorded when AC current densities are applied. This indicates that AC reduces the

CP effectiveness for corrosion protection and induces AC corrosion. Moreover, the anodic current

density is larger with the increasing AC current density.

0 500 1000 1500 2000 2500-1

0

1

2

3

4

CP+AC

CP-IICP-I

Cu

rre

nt

de

nsity (

A/m

2)

Time (s)

10 A/m2

50 A/m2

100 A/m2

200 A/m2

Fig. 9.5. CP current densities of the steel electrode at various AC current densities under

CP potential of -0.850 V (CCS) in NS4 solution.

At the CP level of -0.925 V (CCS), in the CP + AC region, negative current densities are

recorded when applying small AC current densities such as 10 and 50 A/m2. Thus, a more negative

CP potential is able to protect the steel from corrosion under small AC current densities. However,

164

the current density becomes anodic when the AC current density is up 100 A/m2. Obviously, at

high AC current densities, the CP level at -0.925 V (CCS) is not able to provide corrosion

protection to the steel.

0 500 1000 1500 2000 2500-1

0

1

2

3

4

CP-II

CP+AC

CP-I

Cu

rre

nt

de

nsity (

A/m

2)

Time (s)

10A/m2

50 A/m2

100 A/m2

200 A/m2



When the CP potential is further decreased to -1.0 V (CCS), in the CP + AC region, negative

current densities are recorded even when the AC current density of 200 A/m2 is applied on the

steel electrode. Thus, this CP level can protect the steel from corrosion at high AC current

densities. It is noted that the cathodic current density increases with the increasing AC current

density, which is different from the observations at the CP levels of -0.850 V (CCS) and -0.925 V

(CCS).

165

0 500 1000 1500 2000 2500-4

-3

-2

-1

0

CP-IICP+ACCP-I

Cu

rre

nt

de

nsity (

A/m

2)

Time (s)

10A/m2

50 A/m2

100 A/m2

200 A/m2



9.3.3 Weight-loss measurements

Fig. 9.8 shows the corrosion rate of the steel as a function of AC current density and CP

potential after 10 days of weight-loss testing in NS4 solution. Generally, the corrosion rate of the

steel increases with the AC current density. In the absence of CP application, the corrosion rate

can be up to 0.59 mm/y at AC current density of 200 A/m2. At individual AC current density, the

corrosion rate reduces with the negative decrease of the CP potential level. At the CP potential of

-0.850 V (CCS), while the corrosion rate decreases, it is still higher than 0.1 mm/y even at a small

AC current density of 10 A/m2, and is up to 0.34 mm/y with the increase of AC current density to

200 A/m2. With the further decrease of CP level to -0.925 V (CCS) and -1.0 V (CCS), the corrosion

rate further decreases at individual AC current density. Particularly, the corrosion rate of the steel

166

is less than 0.1 mm/y at all AC current densities when the CP potential is as negative as -1.0 V

(CCS).

0 50 100 150 2000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Co

rro

sio

n r

ate

(m

m/y

)


Corrosion potential

-0.850 V (CCS)

-0.925 V (CCS)

-1.0 V (CCS)

Fig. 9.8. Corrosion rate of the steel as a function of AC current density and CP potential

after 10 days of weight-loss testing in NS4 solution.

9.4 Discussion

9.4.1 Effect of AC on shift of CP potential

This chapter demonstrates that the CP potential applied on the steel for corrosion protection

cannot be maintained in the presence of AC interference, and the shift of CP potential, i.e., the

‘true’ DC potential of the steel, depends on the CP potential level and AC current density.

Specifically, the DC potential of the steel is shifted negatively by AC under the applied CP of -

0.850 V (CCS), but positively shifted by AC under the CP of – 1.0 V (CCS). When the CP level

167

is -0.925 V (CCS), the DC potential becomes more positive than the CP potential at small AC

current densities of 10 and 50 A/m2. With the increase of AC current density to 100 A/m2, the DC

potential is shifted negatively.

A mechanistic model, as shown in Fig. 9.9, is proposed to illustrate the effect of AC on DC

potential of steels under various CP levels. At the CP potential of -0.850 V (CCS), the steel is

cathodically polarized and cation ions are adsorbed on steel surface, where the double-charge layer

serves as a current reservoir for Faradic processes. Upon application of AC, excessive ferrous ions

are generated during the positive cycles due to Fe oxidation. The cation ions, i.e., ferrous ions,

flow towards the solution for formation of corrosion products. As a result, the quantity of cation

ions reserved in the double-charger layer reduces, causing the negative shift of DC potential. With

the increase of the AC current density, more ferrous ions are generated during the positive cycles

of AC, and diffuse towards the solution in the electric field, resulting in further negative shifts of

the DC potential (as seen in Fig. 9.2). When a more negative CP potential, such as -0.925 V (CCS),

is applied, the steel is further cathodically polarized, and some ferrous ions are adsorbed and

present in the double-charge layer. This would shift the DC potential of the steel positively if the

AC current density is not sufficiently high, such as 10 and 50 A/m2. When the AC current density

is sufficiently high, the cathodic polarization of the steel at this CP potential becomes not capable

of adsorbing additional ferrous ions which are generated during positive cycles of the AC. The

cation ions would leave the electrode surface and diffuse towards the solution. The leakage of

cation ions results in the negative shift of the DC potential at high AC current densities such as

those over 100 A/m2. For the same reason, when a more negative CP potential of -1.0 V (CCS) is

applied, the steel is polarized cathodically enough to have ferrous ions, which are generated due

to AC corrosion, adsorbed on the steel surface, resulting in a positive shift of the DC potential.

168

With the increase of AC current density, the enhanced AC induced electric field would accelerate

diffusion of cations ions, i.e., ferrous ions, out of the double-charge layer. An equilibrium between

the adsorption of cation ions on the cathodically polarized steel and the diffusion of cation ions

towards the solution can be developed, even at high AC current density such as 200 A/m2.

Furthermore, when CP is stopped, the DC potential of the steel is shifted positively by AC. This

is primarily due to the formation of corrosion products driven by AC corrosion and the deposit on

the steel surface. With the increase in AC current density, the corrosion product layer thickens and

the DC potential is shifted positively since more cation ions are stayed in the double-charge layer.

From the analysis regarding the effect of AC on DC potential of the steel which is under CP,

it is seen that, no matter if the DC potential is shifted negatively or positively upon application of

AC, the enhanced dissolution of the steel by AC and the generation of cation ions, i.e., ferrous

ions, are the common result at the three CP levels testing in this work. The different shifts at

specific CP/AC combinations are resulted from the cathodic polarization of the steel, accelerated

steel corrosion by AC, diffusion of the generated ferrous ions, and the deposit of corrosion products

on the steel surface. In other words, at the three CP levels, the DC potential of the steel would

deviate from the CP potential either negatively or positively, but the steel suffers from increased

corrosion by AC. This conclusion is supported by the continuous increase of the corrosion rate of

the steel as a function of AC current density and CP potential in Fig. 9.8. Therefore, the recorded

DC potential, if it is shifted negatively, in the presence of AC could give misleading information

that the cathodic polarization of the steel is further increased. Actually, the corrosion of the steel

is increased by AC.

169

Fig. 9.9. Schematic diagram of the mechanistic model to illustrate the effect of AC on shift

of the DC potential and at CP potentials (a) -0.850 V (CCS); (b) -0.925 V (CCS); (c) -1.0 V

(CCS).

170

9.4.2 Effect of AC on CP performance

When CP is applied potentiostatically on the steel, the CP current density depends on both

applied CP potential and AC current density. At the CP potential of -0.850 V (CCS), the AC

application results in generation of anodic current densities in CP + AC region (Fig. 9.5), indicating

that the applied CP is not able to cathodically polarize the steel, and protects it from corrosion in

the presence of AC interference, even at 10 A/m2 of AC current density. The larger the AC current

density, the higher the anodic current density is measured, which is associated with increased

corrosion of the steel. When the CP potential of -0.925 V (CCS) is applied, the enhanced cathodic

polarization can protect the steel from corrosion at small AC current densities of 10 and 50 A/m2,

as indicated by the cathodic current densities as recorded. With the increase of AC current density,

the corrosion of the steel is increased, which cannot be protected by the CP, where the anodic

current density is measured. At a more negative CP potential of -1.0 V (CCS), the sufficient

cathodic polarization of the steel is able to prevent AC induced corrosion even at high AC current

densities.

Obviously, the effectiveness of CP for corrosion protection depends on the CP potential

applied and the AC current density, as seen in Fig. 9.8. The industry CP standard of -0.850 V

(CCS) cannot protect steel pipelines under the AC interference, even at low AC current density

such as 10 A/m2. A negative increase of the CP potential is beneficial to protect the AC corrosion.

At the CP level of -0.925 V (CCS), the steel can be protected under small AC current densities of

10 and 50 A/m2 only. When the CP potential is sufficiently negative to -1.0 V (CCS), the steel can

be fully protected against AC corrosion, and the corrosion rate is below an acceptable level of 0.1

mm/y even when AC current density is up to 200 A/m2. Therefore, the AC can decrease the

171

effectiveness of CP for corrosion protection. The CP standard that does not consider the AC

interference is not appropriate for AC corrosion protection.

9.5 Summary

The CP potential applied on steels for corrosion protection cannot be maintained in the

presence of AC interference. The shift of CP potential, i.e., the ‘true’ DC potential of the steel,

depends on the CP potential level and AC current density. The DC potential of the steel is shifted

negatively by AC under the applied CP of -0.850 V (CCS), but positively shifted by AC under the

CP of – 1.0 V (CCS). When the CP level is -0.925 V (CCS), the DC potential becomes more

positive than the CP potential at small AC current densities of 10 and 50 A/m2. With the increase

of AC current density to 100 A/m2, the DC potential is shifted negatively.

The shift of DC potential by AC is resulted from the cathodic polarization level of the steel

under different CP potentials, generation of cation ions such as ferrous ions by corrosion of the

steel during positive cycle of AC signals, mass transfer of the cation ions towards solution, and the

deposit of corrosion product layer on the steel surface. A mechanistic model is developed to

illustrate the dependence of the DC potential shift on CP and AC.

No matter if the DC potential of the steel is shifted negatively or positively upon application

of AC, the steel suffers from increased corrosion. It is thus noted that the recorded DC potential,

if it is shifted negatively, in the presence of AC could give misleading information that the cathodic

polarization of the steel is further increased. Actually, the corrosion of the steel is increased by

AC.

The AC decreases the effectiveness of CP for corrosion protection. At CP potential of -0.850

V (CCS), the applied CP is not able to cathodically polarize the steel, and protects it from corrosion

172

in the presence of AC interference, even at 10 A/m2 of AC current density. When the CP potential

of -0.925 V (CCS) is applied, the enhanced cathodic polarization can protect the steel from

corrosion at small AC current densities of 10 and 50 A/m2. With the increase of AC current density,

the corrosion of the steel is increased, which cannot be protected by the CP. At a more negative

CP potential of -1.0 V (CCS), the sufficient cathodic polarization of the steel is able to prevent AC

induced corrosion even at high AC current densities. The CP standard that does not consider the

AC interference is not appropriate for AC corrosion protection.

173

Chapter Ten: Conclusions and Recommendations

10.1 Conclusions

Through both experimental testing and corrosion modeling, the fundamental aspects of

pipeline external corrosion, coating degradation and CP effectiveness in the presence of AC

interference have been investigated. A sound science foundation has been developed based on the

research outcomes in this work. Primary conclusions are drawn as shown below.

The size and geometric shape of coating defects are critical to AC corrosion of pipelines. The

smaller the coating defect, the larger the AC corrosion rate and the higher probability the AC

induced pitting corrosion, which is due to the larger AC current density generated at the smaller

defect. Compared with circular and square defects, triangular ones are associated with the least

negative DC potential and the smallest anodic dissolution current density. This is attributed to the

sharp corners of the defect that favor accumulation of corrosion products and limit their diffusion.

The corrosion is reduced due to the blocking effect.

In high pH carbonate/bicarbonate solutions, the critical AC current density to initiate pitting

corrosion is approximately 300 A/m2. The passive film, which forms on the steel surface, is

degraded due to the AC induced cathodic polarization and accelerated dissolution of the steel,

resulting in formation of corrosion pits locally. In neutral pH bicarbonate solutions, the threshold

AC current density to initiate pitting is approximately 200 A/m2. The corrosion product layer

generated by active dissolution of the steel can be a barrier to prevent cation ions from diffusing

towards the solution. Pits are generated under the porous corrosion products.

The AC can enhance disbondment of coatings from the steel substrate, increase the water

uptake by the coatings, and decrease the coating resistance. Under AC interference, the epoxy

174

coatings become more loose and porous. Moreover, the AC is able to result in structural changes

of the coatings by generating an electric field on the coating to polarize the functional groups and

enhancing the absorption of water molecules and ions.

The applied CP can be shielded by coating disbondment, and the shielding effect is more

apparent with the increasing disbonding depth. The geometric factor of the coating disbondment

plays an essential role in CP shielding. When the disbondment becomes wider, the shielding effect

is mitigated. The CP shielding can result in separated anodic and cathodic reactions, which occur

at the disbondment bottom and the open holiday, respectively. This is the key mechanism resulting

in localized corrosion under disbonded coating on a cathodically protected pipeline.

The HDPE is a CP shielding coating, which blocks CP current to reach the steel for corrosion

protection when the coating is disbonded and a corrosive environment is developed under the

coating. The FBE is permeable to CP, and CP can penetrate the coating to protect steel from

corrosion under disbonded coating. As confirmed by FT-IR characterizations, after CP permeating

testing, the chemical structure of HDPE does not change. Conversely, there are obvious changes

of the absorbance intensities of functional groups of FBE under CP application. Moreover, the CP

permeation through the FBE coating is time dependent. Over the testing period in this work, (i.e,

after 30 days of testing) the CP is partially permeable through the coating. In other words, the steel

under the disbonded FBE coating is not under full protection. A more negative CP potential level

of -1.0 V (CCS), enhances the CP permeation. The CP permeation through the FBE coating occurs

via three stages: bonding of water molecules with the polar species in the coating, accumulation

of water molecules at accessible micropores existing in the coating, and permeation of water and

ions through the coating via the micropore channels driven by electroendosmosis.

175

The presence of AC interference could affect the CP permeation under the coating

disbondment. At small AC current densities, such as 100 A/m2, the AC enhanced corrosion

generates ferrous and hydroxyl ions, improving the conductivity of solution trapped under the

disbonded coating. This results in enhancement of permeation of CP current into the crevice.

However, with the increase of AC current density, corrosion product generates and deposits in the

solution, blocking the ionic diffusion and CP permeation. When the AC current density is up to

500 A/m2, the CP is blocked completely to reach the disbonding crevice. Thus, it is important to

note that, under high AC current densities, the applied CP can be shielded completely from

reaching the coating disbondment.

The effect of AC on the CP performance depends on the CP potential and the AC current

density applied on the electrode. The CP potential of -0.850 V (CCS) could not protect the steel

under the AC interference, and the measured anodic current density increases with the AC current

density. When CP is -0.925 V (CCS), the steel is partly protected, depending on the applied AC

current density. When the CP potential is sufficiently negative to -1.0 V (CCS), the steel is under

a complete cathodic protection even when the AC current density is up to 200 A/m2.

10.2 Recommendations

This research advances our understanding about AC corrosion of pipelines, and the CP

performance for corrosion protection under AC interference. Further work should be conducted in

order to improve the applicability of the research outcomes in practice. In particular,

1) Application of the home-developed DAQ system in the field for online monitoring AC

voltage and AC current density on pipelines, and assessing the risk of the pipeline to AC

176

corrosion. Moreover, the DAQ system and the relevant results in this work would guide

the industry to evaluate the effectiveness of CP applied on pipelines.

2) While mechanistic studies of pipeline AC corrosion have made big accomplishments in

Cheng group at the University of Calgary, further research can be switched to simulating

and predicting the AC interference based on the information such as orientation between

the AC power lines and the collocated pipelines, soil resistivity, coating status, etc. The

modeling and predictive technology can be applicable for industry use.

177

References

[1] M. Mohitpour, H. Golshan, A. Murray, Transient Flow in Liquid and Gas Pipelines, ASME

press, 2007.

[2] Canadian Energy Pipeline Association (CEPA) website: www.cepa.com.

[3] H.A. Kishawy, H.A. Gabbar, Review of pipeline integrity management practices, Int. J. Pres.

Ves. Pip., 87 (2010) 373-380.

[4] B. McCollum, G. Ahlborn,The influence of frequency of alternating or infrequently reversed

current on electrolytic corrosion, Amer. Inst. Elec. Eng., 35 (1916) 301-345.

[5] U. Bertocci, AC induced corrosion. The effect of an alternating voltage on electrodes under

charge-transfer control, Corrosion, 35 (1979) 211-215.

[6] M.A. Pagano, S.B. Lalvani, Corrosion of mild steel subjected to alternating voltages in

seawater, Corros. Sci., 36 (1994) 127-140.

[7] R. Floyd, Testing and mitigation of AC corrosion on 8 line: a field study, CORROSION'04,

NACE, New Orleans, 2004, paper no. 04210.

[8] Z. Panossian, E.W. Laurino, S.E. Abud Filho, J.H. de L.Oliver, N.L. de Almeida, G. de

S.Pimenta, M.L. P.Filho, J.Á. de C.Albertini, D. de L.Silva, Evaluation of AC corrosion in

pipelines buried using AC probe, in: CORROSION'11, paper no. 11321, NACE, Houston,

TX, 2011.

[9] R. Ellis, Ac induced corrosion on onshore pipelines, a case history, UKOPA publication,

UKOPA-United Kingdom Onshore Pipeline Operators’ Association, (2001).

[10] R.G. Wakelin, C. Sheldon, Investigation and mitigation of AC corrosion on a 300 MM natural

Gas Pipeline, in: Corrosion'04, paper no. 04205, NACE, New Orleans, LO, 2004.

http://www.cepa.com/

178

[11] D.-K. Kim, S. Muralidharan, T.-H. Ha, J.-H. Bae, Y.-C. Ha, H.-G. Lee, J.D. Scantlebury,

Electrochemical studies on the alternating current corrosion of mild steel under cathodic

protection condition in marine environments, Electrochim. Acta, 51 (2006) 5259-5267.

[12] F. Kajiyama, Y. Nakamura, Effect of induced alternating current voltage on cathodically

protected pipelines paralleling electric power transmission lines, Corrosion, 55 (1999) 200-

205.

[13] L.V. Nielsen, F. Galsgaard, Sensor Technology for On-Line Monitoring of AC Induced

Corrosion Along Pipelines, in: Corrosion'05, paper no. 05375, NACE, Houston, TX, 2005.

[14] R.A. Gummow, R.G. Wakelin, S.M. Segall, AC corrosion--A new challenge to pipeline

integrity, in, Corrosion'1998, paper no. 98566, Houston, TX, NACE 1998.

[15] M. Ormellese, L. Lazzari, S. Goidanich, V. Sesia, CP criteria assessment in the presence of

AC interference, in: CORROSION 2008, paper no. 08064, NACE, New Orleans, LO, 2008.

[16] L.V. Nielsen, Role of Alkalization in AC Induced Corrosion of Pipelines and Consequences

Hereof in Relation to CP Requirements, in: Corrosion'2005, paper no. 05188, NACE,

Houston, TX, 2005.

[17] L.Y. Xu, X. Su, Y.F. Cheng, Effect of alternating current on cathodic protection on pipelines,

Corros. Sci., 66 (2013) 263-268.

[18] Z. Jiang, Y. Du, M. Lu, Y. Zhang, D. Tang, L. Dong, New findings on the factors accelerating

AC corrosion of buried pipeline, Corros. Sci., 81 (2014) 1-10.

[19] L.V. Nielsen, K. Nielsen, B. Baumgarten, H. Breuning-Madsen, P. Cohn, H. Rosenberg, AC

induced corrosion in pipelines: detection, characterization and mitigation, in: CORROSION

2004, paper no. 04211, NACE, New Orleans, LO, 2004.

179

[20] L. Bertolini, M. Carsana, P. Pedeferri, Corrosion behaviour of steel in concrete in the presence

of stray current, Corros. Sci., 49 (2007) 1056-1068.

[21] L. Di Biase, Interaction and stray-current corrosion, Shreir's Corrosion, Vol. 4, Chapter 4.22,

pp. 2833-2838, Elsevier, 2010.

[22] Y. Li, X. Li, G. Cai, L. Yang, Influence of AC interference to corrosion of Q235 carbon steel,

Corros. Eng. Sci. Technol., 48 (2013) 322-326.

[23] H.R. Hanson, J. Smart, AC Corrosion on a Pipeline Located in a HVAC Utility Corridor, in:

Corrosion'2004, paper no. 04209, NACE, New Orleans, LO.

[24] W.H. Bruckner, The effects of 60 cycle alternating current on the corrosion of steels and other

metals buried in soils [R]. Illinois: Engineering Experiment Station Bulletion, University of

Illinois, Urbana, 1964.

[25] R.G. Wakelin, R.A. Gummow, S.M. Segall, AC Corrosion - Case Histories, Test Procedures,

& Mitigation, in: Corrosion'1998, paper no. 98565, NACE, San Diego, CA, 1998.

[26] D. Funk, H.G. Schoeniech, Problems with coupons when assessing the AC-corrosion risk of

pipelines, Spec. Steel Pipe., 41 (2002) 54.

[27] L. Lazzari, P. Pedeferri, Cathodic protection, pp. 370, ISBN: 8873980201, Polipress, 2006. .

[28] K. Yin, M. Tang, C. Hsiong, Corrosion of buried pipeline under effect of electrical field with

industrial frequency, International Petroleum Exhibition and Technical Symposium, Society

of Petroleum Engineers, 1982.

[29] NACE SP0177, Mitigation of alternating current and lightning effects on metallic structures

and corrosion control systems, NACE International Standard Practice, 2007.

[30] R. W. Bonds, The effect of overhead AC power lines paralleling ductile iron pipelines, DIPRA

publication, DIPRA - Ductile Iron Pipe Research Association, 1997.

180

[31] S. Finneran, B. Krebs, Advances in HVAC Transmission Industry and its Effects on Pipeline

Induced AC Corrosion, in: Corrosion'2014, paper no. 4421, NACE, San Antonio, TX, 2014.

[32] L. Bortels, C. Baeté, J.-M. Dewilde, Accurate modeling and troubleshooting of ac interference

problems on pipelines, in: CORROSION'12, paper no. 1664, NACE, Salt Lake City, UT,

2012.

[33] L. Qi, H. Yuan, Y. Wu, X. Cui, Calculation of overvoltage on nearby underground metal

pipeline due to the lightning strike on UHV AC transmission line tower, Electr. Pow. Syst.

Res., 94 (2013) 54-63.

[34] C. Movley, Pipeline Corrosion From Induced A.C, in: CORROSION'05, paper no. 05132,

NACE, Houston, TX, 2005.

[35] L.V. Nielsen, P. Cohn, AC corrosion and electrical equivalent diagrams, CeoCor, Bruxelles,

Belgium, 2000.

[36] I. Ragault, AC Corrosion Induced by V.H.V. Electrical Lines on Polyethylene Coated Steel

Gas Pipelines, in: Corrosion'1998, paper no. 98557, NACE, San Diego, CA, 1998.

[37] J. Williams, Corrosion of metals under the influence of alternaitng current, Mater. Prot., 5

(1996) 52.

[38] M. Ormellese, S. Goidanich, F. Bolzoni, L. Lazzari, M.P. Pedeferri, Laboratory testing on the

influence of alternated current on steel corrosion, CORROSION'04, NACE, 2004, paper no.

04208.

[39] P. Linhardt, G. Ball, AC corrosion: results from laboratory investigations and from a failure

analysis, CORROSION'06, NACE, Houston, 2006, paper no. 06160.

181

[40] M. Yunovich, N.G. Thompson, AC corrosion: mechanism and proposed model, in: 2004

International Pipeline Conference, American Society of Mechanical Engineers, 2004, pp.

183-195.

[41] M. Buchler, H. Schoneich, Investigation of alternating current corrosion of cathodically

protected pipelines: development of a detection method, mitigation measures, and a model

for the mechanism, Corrosion, 65 (2009) 578-586.

[42] L.V. Nielsen, Role of alkalization in AC induced corrosion of pipelines and consequences

hereof in realtion to CP requirements, CORROSION'05, NACE, Houston, 2005, paper no.

05188.

[43] A. Brenna, M. Ormellese, L. Lazzari, A Proposal of AC Corrosion Mechanism of Carbon

Steel in Cathodic Protection Condition, in: Corrosion'2013, paper no. 2457, NACE Orlando,

FL, 2013.

[44] Z. Panossian, S.E. Filho, N.L. de Almeida, M.L. Pereira Filho, D. de L.Silva, E.W. Laurino,

J.H. L.Oliver, G. de S.Pimenta, J.Á. C.Albertini, Effect Of Alternating Current By High

Power Lines Voltage And Electric Transmission Systems In Pipelines Corrosion, in:

Corrosion'2009, paper no. 09541, NACE, Atlanta, GE, 2009.

[45] S. Lalvani, X. Lin, A revised model for predicting corrosion of materials induced by

alternating voltages, Corros. Sci., 38 (1996) 1709-1719.

[46] L.Y. Xu, X. Su, Z.X. Yin, Y.H. Tang, Y.F. Cheng, Development of a real-time AC/DC data

acquisition technique for studies of AC corrosion of pipelines, Corros. Sci., 61 (2012) 215-

223.

[47] S. Lalvani, G. Zhang, The corrosion of carbon steel in a chloride environment due to periodic

voltage modulation: Part I, Corros. Sci., 37 (1995) 1567-1582.

182

[48] C. Wen, J. Li, S. Wang, Y. Yang, Experimental study on stray current corrosion of coated

pipeline steel, Journal of Natural Gas Science and Engineering, 27, Part 3 (2015) 1555-1561.

[49] A. Fu, Y. Cheng, Effects of alternating current on corrosion of a coated pipeline steel in a

chloride-containing carbonate/bicarbonate solution, Corros. Sci., 52 (2010) 612-619.

[50] S. Lalvani, G. Zhang, The corrosion of carbon steel in a chloride environment due to periodic

voltage modulation: part II, Corros. Sci., 37 (1995) 1583-1598.

[51] S. Goidanich, L. Lazzari, M. Ormellese, AC corrosion–Part 1: Effects on overpotentials of

anodic and cathodic processes, Corros. Sci., 52 (2010) 491-497.

[52] H. Xiao, S.B. Lalvani, A linear model of alternating voltage-induced corrosion, J. Electro.

Soc., 155 (2008) C69-C74.

[53] D.Z. Tang, Y.X. Du, M.X. Lu, Y. Liang, Z.T. Jiang, L. Dong, Effect of pH value on corrosion

of carbon steel under an applied alternating current, Materials and Corrosion, 66 (2015) 1467-

1479.

[54] D. Tang, Y. Du, M. Lu, G. Wu, Y. Liang, Z. Jiang, Effect of Alternating Current on Corrosion

of Pipeline Steel, in: CORROSION'15, paper no. 5626, NACE, Dallas, TX, 2015.

[55] S. Finneran, B. Krebs, Advances in HVAC Transmission Industry and its Effects on Pipeline

Induced AC Corrosion, in: CORROSION'14, paper no. 4421, NACE, San Antonio, TX,

2014.

[56] K. Yin, M. Tang, C. Hsiong, Corrosion of Buried Pipeline Under Effect of Electrical Field

With Industrial Frequency, in, Soc. Petro. Eng, SPE-10578, Beijing, China, 1982.

[57] K.V. Quang, F. Brindel, G. Laslaz, R. Buttoudin, Pitting mechanism of aluminum in

hydrochloric acid under alternating current, J. Electrochem. Soc., 130 (1983) 1248-1252.

183

[58] R. Ruedisueli, H. Hager, C. Sandwith, An application of a state-of-the-art corrosion

measurement system to a study of the effects of alternating current on corrosion, Corrosion,

43 (1987) 331-338.

[59] Y.-B. Guo, C. Liu, D.-G. Wang, S.-H. Liu, Effects of alternating current interference on

corrosion of X60 pipeline steel, Pet. Sci., 12 (2015) 316-324.

[60] S. Fernandes, S. Mehendale, S. Venkatachalam, Influence of frequency of alternating current

on the electrochemical dissolution of mild steel and nickel, J. Appl. Electrochem., 10 (1980)

649-654.

[61] W.W. Qiu, M. Pagano, G. Zhang, S. Lalvani, A periodic voltage modulation effect on the

corrosion of Cu-Ni alloy, Corros. Sci., 37 (1995) 97-110.

[62] CEN/TS 15280, Evaluation of AC corrosion likelihood of buried pipelines - application on

cathodically protected pipeliens, London, England: BSI.

[63] D.E. Gilroy, AC Interference-Important Issues for Cross Country Pipelines, in:

CORROSION 2003, paper no. 03699, NACE, San Diego, CA, 2003, 2003.

[64] D.A. Eden, AC interference and accelerated pipeline corrosion in utility corridors–an

alternative perspective to the root causes, Corrosion/2006 paper no. 06159, Houston, TX,

2006, NACE, (2006).

[65] G. Helm, T. Helm, H. Heinzen, W. Schwenk, Investigation of corrosion of cathodically

protected steel subjected to alternating currents, Corros. Sci., 5 (1993): 246.

[66] DIN 50925, Corrosion of metals, proof of effectiveness of cathodic corrosion protection of

underground installations, Berlin, Germany.

[67] W.V. Baeckmann, W. Schwenk, Handbuch des Kathodischen Korrosionsschutzes,

Weinheim, Germany: Wiley-VCH, 1999.

184

[68] M. Yunovich, N.G. Thompson, AC Corrosion: Corrosion Rate and Mitigation Requirements,

in: Corrosion'2004, paper no. 04206, NACE, New Orleans, LO.

[69] S. Goidanich, L. Lazzari, M. Ormellese, AC corrosion. Part 2: Parameters influencing

corrosion rate, Corros. Sci., 52 (2010) 916-922.

[70] M.A. Pagano, S.B. Lalvani, Corrosion of mild steel subjected to alternating voltages in

seawater, Corrosion science, 36 (1994) 127-140.

[71] E. Collet, B. Delores, M. Gabillard, I. Ragault, Corrosion due to AC influence of very high

voltage power lines on polyethylene-coated steel pipelines: evaluation of risks-preventive

measures, Anti-Corros. Method. M., 48 (2001) 221-227.

[72] R. Norsworthy, Addressing Soil Stress and CP Shielding by Using a Woven Geotextile Fabric

Backed Tape System With a Rubber Modified Bituminous Compound, in: Corrosion'2000,

paper no. 00769, NACE, Orlando, FL, 2000.

[73] D. Funk, W. Prinz, H. Schoneich, Investigations of AC corrosion in cathodically protected

pipes, Ochrona Przed Korozja, 36 (1993) 225-228.

[74] NACE International Task Group 327. Ac corrosion state-of-the-art: Corrosion rate,

mechanism, and mitigation requirements. NACE International. Publication 35110, (24242),

2010.

[75] SP0169-2013 formerly RP0169 Control of External Corrosion on Underground or Submerged

Metallic Piping Systems, NACE Internation Standard, 2013.

[76] L.S. Lilleby, S.M. Hesjevik, S. Olsen, Effects From Alternating Current On Cathodic

Protection Of Submarine Pipelines, in: CORROSION'11, paper no. 11055, NACE, Houston,

TX, 2011.

185

[77] K.N. Allahar, M.E. Orazem, On the extension of CP models to address cathodic protection

under a delaminated coating, Corros. Sci., 51 (2009) 962-970.

[78] M.E. Parker, E.G. Peattie, Pipeline Corrosion and Cathodic Protection, third ed., Gulf

Professional Publishing, USA, 1999.

[79] W. von Baeckmann, W. Schwenk, W. Prinz, Handbook of cathodic corrosion protection, Gulf

Professional Publishing, 1997.

[80] M. Büchler, C. Voute, F. Stalder, Characteristics of potential measurements in the field of AC

corrosion, Ceocorr, Title 15, Sector A. 2003, (2003).

[81] A. Eslami, R. Kania, B. Worthingham, G. Boven, R. Eadie, W. Chen, Corrosion of X-65

pipeline steel under a simulated cathodic protection shielding coating disbondment,

Corrosion, 69 (2013) 1103-1110.

[82] Y. Hosokawa, F. Kajiyama, Y. Nakamura, Overcoming the New Threat to Pipeline Integrity-

AC corrosion assessment and its mitigation, in: Proc. 23rd World Gas Conference,

Amsterdam, Holland, 2006.

[83] D. Tang, M. Lu, Y. Du, Z. Jiang, L. Dong, Study on CP Criteria for Mild Steel in the Presence

of AC, in: Corrosion'2014, paper no. 3802, NACE, San Antonio, TX, 2014.

[84] L.J. Zhang, Y.W. Pan, G. Tao, Y.M. Wang, Y. Li, Study of Effect and Protection for

Corrosion on Buried Pipeline by AC Interference, in: Advanced Materials Research, Trans

Tech Publ, 2012, pp. 1756-1763.

[85] S. Goidanich, L. Lazzari, M. Ormellese, M.P. Pedeferri, Effect of AC on cathodic protection

of carbon steel in simulated soil conditions, EUROCORR 2006, paper no. 279.

186

[86] H.-S. Song, Y.-T. Kho, Y.-G. Kim, S.-M. Lee, Y.S. Park, Competition of AC and DC Current

in AC Corrosion Under Cathodic Protection, in: Corrosion'2002, paper no. 02117, NACE,

Denver, CO, 2002.

[87] A.Q. Fu, Y.F. Cheng, Effect of alternating current on corrosion and effectiveness of cathodic

protection of pipelines, Can. Metall. Q., 51 (2012) 81-90.

[88] Y. Hosokawa, F. Kajiyama, Y. Nakamura, New CP criteria for elimination of the risks of AC

corrosion and overprotection on cathodically protected pipelines, in: CORROSION 2002,

paper no. 02111, NACE, Denver, CO, 2002.

[89] P. Carpentiers, R. Gregoor, A. Pourbaix, AC corrosion: detection, investigations and

mechanisms, EUROCORR 2003, paper no. 307.

[90] M. Ormellese, L. Lazzari, A. Brenna, A. Trombetta, Proposal Of Cp Criterion In The Presence

Of Ac-Interference, in: CORROSION'10, paper no. 10032, NACE, San Antonio, TX, 2010.

[91] M. Ormellese, S. Goidanich, L. Lazzari, Effect of AC interference on cathodic protection

monitoring, Corros. Engi., Sci. Technol., 46 (2011) 618-623.

[92] G. Camitz, C. Johansson, A. Marbe, Alternating current corrosion on cathodically protected

steel in soil - A long-term field investigation. CeoCor (Committee on the Study of Pipe

Corrosion and Protection) International Congress, Bruxelles, Belgium, 2000. .

[93] R.G. Wakelin, R.A. Gummow, S.M. Segall, AC corrosion-case histories, test procedures, &

mitigation, CORROSION 98, (1998).

[94] I. Ragault, AC corrosion induced by VHV electrical lines on polyethylene coated steel gas

pipelines, CORROSION 98, (1998).

[95] S. Goidanich, L. Lazzari, M. Ormellese, AC corrosion - Part 1: Effects on overpotentials of

anodic and cathodic processes, Corrosion science, 52 (2010) 491-497.

187

[96] M. Ormellese, A. Brenna, L. Lazzari, Effects of AC-interference On Passive Metals

Corrosion, in: CORROSION'2011, paper no. 11342, NACE, Houston, TX, 2011.

[97] Y. Li, X. Li, G. Cai, L. Yang, Influence of AC interference to corrosion of Q235 carbon steel,

Corr. Eng. Sci. Technol., 48 (2013) 322-326.

[98] L. Xu, X. Su, Y. Cheng, Effect of alternating current on cathodic protection on pipelines,

Corr. Sci., 66 (2013) 263-268.

[99] L. Xu, X. Su, Z. Yin, Y. Tang, Y. Cheng, Development of a real-time AC/DC data acquisition

technique for studies of AC corrosion of pipelines, Corr. Sci., 61 (2012) 215-223.

[100] A. Fu, Y. Cheng, Effect of alternating current on corrosion and effectiveness of cathodic

protection of pipelines, Canadian Metallurgical Quarterly, 51 (2012) 81-90.

[101] C. Dong, A. Fu, X. Li, Y. Cheng, Localized EIS characterization of corrosion of steel at

coating defect under cathodic protection, Electrochimica Acta, 54 (2008) 628-633.

[102] X. Xu, L. Xu, W. Liu, M. Lu, Study of AC Corrosion On Coated Pipeline Steel In Simulated

Soil Solution, in: CORROSION'2011, paper no. 11343, NACE, Houston, TX, 2011.

[103] Z. Jiang, Y. Du, M. Lu, Y. Zhang, D. Tang, L. Dong, New findings on the factors

accelerating AC corrosion of buried pipeline, Corros. Sci., 81 (2013) 1-10.

[104] A.Q. Fu, Y.F. Cheng, Effects of alternating current on corrosion of a coated pipeline steel in

a chloride-containing carbonate/bicarbonate solution, Corrosion science, 52 (2010) 612-619.

[105] G. Zhang, Y. Cheng, On the fundamentals of electrochemical corrosion of X65 steel in CO

2-containing formation water in the presence of acetic acid in petroleum production, Corros.

Sci., 51 (2009) 87-94.

[106] S. Lalvani, X. Lin, A theoretical approach for predicting AC-induced corrosion, Corros. Sci.,

36 (1994) 1039-1046.

188

[107] W. Yongji, L. Xiangyi, Electrochemical Behaviors And SCC Sensitive Potential Of Pipeline

Steels In Carbonate-Bicarbonate Solutions, in: CORROSION'04, paper no. 04201, NACE,

New Orleans, LO, 2004.

[108] B. Ingham, M. Ko, N. Laycock, J. Burnell, P. Kappen, J. Kimpton, D. Williams, In situ

synchrotron X-ray diffraction study of scale formation during CO 2 corrosion of carbon steel

in sodium and magnesium chloride solutions, Corros. Sci., 56 (2012) 96-104.

[109] F. King, T.R. Jack, W. Chen, M.J. Wilmott, R.R. Fessler, K. Krist, Mechanistic Studies of

Initiation and Early Stage Crack Growth for Near-Neutral pH SCC on Pipelines, in:

CORROSION'00, paper no. 00361, NACE, Orlando, FL, 2000.

[110] R. Armstrong, A. Coates, The passivation of iron in carbonate/bicarbonate solutions, J.

Electro. Chem. Interf. Electrochem., 50 (1974) 303-313.

[111] J. Heuer, J. Stubbins, An XPS characterization of FeCO3 films from CO2 corrosion, Corros.

Sci., 41 (1999) 1231-1243.

[112] A. Fu, Y. Cheng, Electrochemical polarization behavior of X70 steel in thin

carbonate/bicarbonate solution layers trapped under a disbonded coating and its implication

on pipeline SCC, Corros. Sci., 52 (2010) 2511-2518.

[113] B. Linter, G. Burstein, Reactions of pipeline steels in carbon dioxide solutions, Corros. Sci.,

41 (1999) 117-139.

[114] G.Z. Meng, C. Zhang, Y.F. Cheng, Effects of corrosion product deposit on the subsequent

cathodic and anodic reactions of X-70 steel in near-neutral pH solution, Corros. Sci., 50

(2008) 3116-3122.

[115] D. Davies, G. Burstein, The effects of bicarbonate on the corrosion and passivation of iron,

Corrosion, 36 (1980) 416-422.

189

[116] J.R. Galvele, Transport processes and the mechanism of pitting of metals, J. Electrochem.

Soc., 123 (1976) 464-474.

[117] C. Chao, L. Lin, D. Macdonald, A point defect model for anodic passive films I. Film growth

kinetics, J. Electrochem. Soc., 128 (1981) 1187-1194.

[118] N. Sato, A theory for breakdown of anodic oxide films on metals, Electrochim. Acta, 16

(1971) 1683-1692.

[119] T. Hoar, The production and breakdown of the passivity of metals, Corros. Sci., 7 (1967)

341-355.

[120] C. Townend, R. Worthington, M. Munro, A comparative review of some modern pipeline

coatings and experience gained applying a polyurethane-tar coating, Anti-Corros. Methods

and Mater., 27 (1980) 5-11.

[121] E.O. Eltai, J. Scantlebury, E. Koroleva, Protective properties of intact unpigmented epoxy

coated mild steel under cathodic protection, Prog. Organ. Coat., 73 (2012) 8-13.

[122] R. Norsworthy, Fusion Bonded Epoxy - A Field Proven Fail Safe Coating System, in:

CORROSION'06, paper no. 06044, NACE, San Diego, CA, 2006.

[123] A. Fu, Y. Cheng, Characterization of the permeability of a high performance composite

coating to cathodic protection and its implications on pipeline integrity, Prog. Org. Coat., 72

(2011) 423-428.

[124] J. Holub, D.T. Wong, M. Tan, Analysis Of Cdt Methods And Factors Affecting Cathodic

Disbondment, in: Corrosion'2007, paper no. 07022, NACE, Nashvile, TE, 2007.

[125] G. Howell, Y. Cheng, Characterization of high performance composite coating for the

northern pipeline application, Prog. Org. Coat., 60 (2007) 148-152.

190

[126] A. Barbucci, M. Delucchi, G. Cerisola, Organic coatings for concrete protection: liquid water

and water vapour permeabilities, Prog. Org. Coat., 30 (1997) 293-297.

[127] F. Mansfeld, C. Tsai, Determination of coating deterioration with EIS: I. Basic relationships,

Corrosion, 47 (1991) 958-963.

[128] K. Chike, M. Myrick, R. Lyon, S. Angel, Raman and near-infrared studies of an epoxy resin,

Appl. Spectro., 47 (1993) 1631-1635.

[129] I. Blanco, G. Cicala, M. Costa, A. Recca, Development of an epoxy system characterized by

low water absorption and high thermomechanical performances, J. Appl. Polym. Sci., 100

(2006) 4880-4887.

[130] P. Musto, L. Mascia, G. Ragosta, G. Scarinzi, P. Villano, The transport of water in a

tetrafunctional epoxy resin by near-infrared Fourier transform spectroscopy, Polymer, 41

(2000) 565-574.

[131] J. Mijovic, S. Andjelic, A study of reaction kinetics by near-infrared spectroscopy. 1.

Comprehensive analysis of a model epoxy/amine system, Macromolecules, 28 (1995) 2787-

2796.

[132] G. George, P. Cole‐Clarke, N. St John, G. Friend, Real‐time monitoring of the cure

reaction of a TGDDM/DDS epoxy resin using fiber optic FT‐IR, J. Appl. Polym. Sci., 42

(1991) 643-657.

[133] M.G. González, J. Baselga, J.C. Cabanelas, Applications of FTIR on epoxy resins-

identification, monitoring the curing process, phase separation and water uptake, INTECH

Open Access Publisher, 2012.

191

[134] K.E. Chike, M.L. Myrick, R.E. Lyon, S.M. Angel, Raman and near-infrared studies of an

epoxy resin, Applied Spectroscopy, 47 (1993) 1631-1635.

[135] E. Mertzel, J.L. Koenig, APPLICATION OF FT-IR AND NMR TO EPOXY RESINS,

Advances in Polymer Science, 75 (1986) 73-112.

[136] S. Meure, D.Y. Wu, S.A. Furman, FTIR study of bonding between a thermoplastic healing

agent and a mendable epoxy resin, Vibrational Spectroscopy, 52 (2010) 10-15.

[137] C.G. Munger, L.D. Vincent, Corrosion prevention by protective coatings, Houston, TX:

NACE International, 1999, (1999).

[138] J. Perdomo, I. Song, Chemical and electrochemical conditions on steel under disbonded

coatings: the effect of applied potential, solution resistivity, crevice thickness and holiday

size, Corros. Sci., 42 (2000) 1389-1415.

[139] D.-T. Chin, G. Sabde, Current distribution and electrochemical environment in a

cathodically protected crevice, Corrosion, 55 (1999) 229-237.

[140] A. Toncre, N. Ahmad, Cathodic protection in crevices under disbonded coatings, A. C.

Toncre and N. Ahmad, Arabian American Oil Company, Dhahran, Saudi Arabia. Mater.

Perform., 19 (1980) 1980.

[141] T. Jack, G. Van Boven, M. Wilmott, R. Sutherby, R. Worthingham, Cathodic protection

potential penetration under disbonded pipeline coating, Mater. Perf., 33 (1994) 17-21.

[142] A. Fu, Y. Cheng, Characterization of corrosion of X65 pipeline steel under disbonded

coating by scanning Kelvin probe, Corros. Sci., 51 (2009) 914-920.

[143] M. Baker Jr, Integrity Management Program—Stress Corrosion Cracking Studies, Final

Report, Office of Pipeline Safety TTO-8, Department of Transportation, (2004).

192

[144] D. Neal, Pipeline Coating Failure - Not Always What You Think It Is, in: CORROSION'00,

paper no. 00755, NACE, Orlando, FL, 2000.

[145] X. Chen, X. Li, C. Du, Y. Cheng, Effect of cathodic protection on corrosion of pipeline steel

under disbonded coating, Corros. Sci., 51 (2009) 2242-2245.

[146] F. Song, N. Sridhar, Modeling pipeline crevice corrosion under a disbonded coating with or

without cathodic protection under transient and steady state conditions, Corros. Sci., 50

(2008) 70-83.

[147] R. Norsworthy, Is your pipeline coating fail/safe, Pipe. Gas J., 233 (2006) 62.

[148] S. Papavinasam, M. Attard, R.W. Revie, External polymeric pipeline coating failure modes,

Mater. Perform., 45 (2006) 28-30.

[149] E. Charles, R. Parkins, Generation of stress corrosion cracking environments at pipeline

surfaces, Corrosion, 51 (1995) 518-527.

[150] J.L. Pikas, Case Histories of External Microbiologically Influenced Corrosion Underneath

Disbonded Coatings, in: CORROSION'96, paper no. 96198, NACE, Denver, CO, 1996.

[151] R. Norsworthy, Coatings Used in Conjuction with Cathodic Protection-Shielding vs. Non-

Shielding Pipeline Coatings, in: 17th international corrosion congress, Las Vegas, Nevada,

paper, 2008.

[152] T. Jack, F. King, Y. Cheng, R. Worthingham, Permeable coatings and CP compatibility, in:

2002 4th International Pipeline Conference, American Society of Mechanical Engineers

New York, 2002, 2002, pp. 1889-1893.

[153] I. Thompson, J.R. Saithala, Review of pipeline coating systems from an operator's

perspective, Corros. Eng. Sci. Technol., 19 (2015) 33.

193

[154] J.A. Beavers, N.G. Thompson, Corrosion Beneath Disbonded Coatings: A Review, in:

Corrosion'96, paper no. 96208, NACE, Denver, CO, 1996.

[155] F. King, T.R. Jack, M. Kolar, R.G. Worthingham, A Permeable Coating Model For

Predicting The Environment At The Pipe Surface Under Cp-Compatible Coatings, in:

CORROSION'04, paper no. 04158, NACE, New Orleans, LO, 2004.

[156] U. Rammelt, G. Reinhard, Application of electrochemical impedance spectroscopy (EIS) for

characterizing the corrosion-protective performance of organic coatings on metals, Prog.

Org. Coat., 21 (1992) 205-226.

[157] M. Yan, C. Sun, J. Xu, T. Wu, S. Yang, W. Ke, Stress corrosion of pipeline steel under

occluded coating disbondment in a red soil environment, Corros. Sci., 93 (2015) 27-38.

[158] G. Ruschau, Y. Chen, Determining the CP Shielding Behavior of Pipeline Coatings in the

Laboratory, in: NACE'06, paper no. 06043, NACE, San Diego, CA, 2006.

[159] R. Norsworthy, S. Brockhaus, J. Grillenberger, M. Ginten, EMAT, Pipe Coatings, Corrosion

Control and Cathodic Protection Shielding, in: CORROSION'13, paper no. 2378, NACE,

Orlando, FL, 2013.

[160] H. Leidheiser, Electrical and electrochemical measurements as predictors of corrosion at the

metal—organic coating interface, Prog. Org. Coat., 7 (1979) 79-104.

[161] M. Orton, The effectiveness of cathodic protection under unbonded coatings on pipelines,

Mater. Perform., 24 (1985) 6.

[162] B. Cherry, A. Gould, Pitting corrosion of nominally protected land-based pipelines, Mater.

Perform., 29 (1990) 8.

[163] S. Mohan, A. Prabakaran, Infrared and Laser Raman Spectra of Polyethylene and its Normal

Coordinate Analysis, Asian J. Chem., 1 (1989) 162-165.

194

[164] S. Krimm, C. Liang, G. Sutherland, Infrared spectra of high polymers. II. Polyethylene, J.

Chem. Phys., 25 (1956) 549-562.

[165] J. Gulmine, P. Janissek, H. Heise, L. Akcelrud, Polyethylene characterization by FTIR,

Polym. Test., 21 (2002) 557-563.

[166] E. Mertzel, J.L. Koenig, Application of FT-IR and NMR to epoxy resins, in: Epoxy Resins

and Composites II, Springer, 1986, pp. 73-112.

[167] S. Meure, D.-Y. Wu, S.A. Furman, FTIR study of bonding between a thermoplastic healing

agent and a mendable epoxy resin, Vib. Spectrosc., 52 (2010) 10-15.

[168] I. Blanco, G. Cicala, M. Costa, A. Recca, Development of an epoxy system characterized by

low water absorption and high thermomechanical performances, Journal of applied polymer

science, 100 (2006) 4880-4887.

[169] N.L. Thomas, The barrier properties of paint coatings, Progress in Organic Coatings, 19

(1991) 101-121.

[170] R. Buchanan, A Critical Review of Industry Codes and Standards as They Relate to

Electrically Resistive Coatings and What That Means to Shielding, CORROSION 2013,

(2013).

[171] M.C. Lee, N.A. Peppas, Water transport in epoxy resins, Prog. Polym. Sci., 18 (1993) 947-

961.

[172] H. Leidheiser Jr, R.D. Granata, Ion transport through protective polymeric coatings exposed

to an aqueous phase, IBM J. Res. Dev., 32 (1988) 582-590.

[173] D. Kuang, Y.F. Cheng, Understand the AC induced pitting corrosion on pipelines in both

high pH and neutral pH carbonate/bicarbonate solutions, Corros. Sci., 85 (2014) 304-310.

195

[174] D.-K. Kim, T.-H. Ha, Y.-C. Ha, J.-H. Bae, H.-G. Lee, D. Gopi, J. Scantlebury, Alternating

current induced corrosion, Corros. Eng. Sci. Technol., 39 (2004) 117-123.

[175] L. Lazzari, S. Goldanich, M. Ormellese, F. Bolzoni, Laboratory Test Results of AC

Interference on Polarized Steel, in: Corrosion'2003, paper no. 03704, NACE, San Diego,

CA, 2003.

[176] D. Kuang, Y.F. Cheng, Study of cathodic protection shielding under coating disbondment

on pipelines, Corros. Sci., 99 (2015) 249-257.

[177] ASTM G 1-03 Standard practice for preparing, cleaning, and evaluating corrosion test

specimens, ASTM International, 2003, pp. 1-9.

pipeline corrosion and coating failure under alternating

Documents