CO

2
andH
2
SCorrosioninOilPipelines

MasterThesis
of
MythiliKoteeswaran

FacultyofMathematicsandNaturalScience
June2010

5
Abstract
Thisstudyhasbeenconductedtofindthecorrosionbehaviorandcorrosionratesof
carbon steel in the presence of CO
2
and H
2
S at various pH levels using classical
electrochemicaltechniques.Itwasfoundthatinagalvaniccoupling,themetalinthe
sulfide environment gets protection even at pH 3, and the bare metal which is in
neutral pH was corroding sacrificially. The linear polarization resistance
measurementsandpotentiodynamicscanofthemetalwithoutthegalvaniccoupling
showahighdegreeofcorrosionatpH3.Thecorrosionrategenerallywashigherfor
CO
2
/H
2
SsystemthanforH
2
Ssystem.

7
Acknowledgement
I would like to express my sincere gratitude to Prof. Tor Hemmingsen for his
continuous academic and moral support. This thesis work is a tribute to his
exceptionalguidanceandmentorship.
I would like to acknowledge my indebtedness to Tor Gulliksen, for helping me
makingthegalvaniccellandthesamples.
IwouldliketoacknowledgemyindebtednesstoOlaRisvikforhelpingmeingetting
theSEMpictures.
IwouldliketothankLivMargarethAkslandforhersupportinthelaboratorywork.
I also would like to acknowledge Koteeswaran Paulpandian for his priceless
suggestionsandrecommendationsinpreparingthethesisreport.

9
TABLEOFCONTENTS

1.INTRODUCTION........................................................................................................15
2.LITERATUREREVIEW................................................................................................16
2.1.CO
2
Corrosion....................................................................................................16
2.1.1TheeffectofpH ..........................................................................................17
2.1.2Theeffectoftemperature ..........................................................................17
2.2.H
2
SCorrosion....................................................................................................18
2.2.1.TheeffectofpH .........................................................................................19
2.2.2.TheeffectofH
2
SConcentration................................................................19
2.2.3.Theeffectoftemperature .........................................................................19
2.3.CO
2
/H
2
SCorrosion ............................................................................................19
2.4.Corrosionproductfilmformation.....................................................................21
2.4.1.Ironcarbide(Fe
3
C) .....................................................................................21
2.4.2.Ironcarbonate(FeCO
3
) ..............................................................................21
2.4.3.Ironsulfide(FeS)film.................................................................................22

3.ELECTROCHEMICALMETHODS ...............................................................................25
3.1GalvanicCorrosion.............................................................................................25
3.2LinearPolarizationresistance............................................................................26
3.2.1Calculationofcorrosionratefromcorrosioncurrent ................................28
3.3Potentiodynamicscan........................................................................................29
3.3.1TheAnodicscan ..........................................................................................29
3.3.2CathodicScan..............................................................................................30
3.3.3CorrosionratefromPotentiodynamicscan................................................31
3.4ElectrochemicalImpedanceSpectroscopy ........................................................32
3.4.1Corrosionratefromimpedanceplot ..........................................................35

4.EXPERIMENTALPROCEDUREANDSETUP................................................................37
5.RESULTSANDDISCUSSION ......................................................................................41
6.CONCLUSION............................................................................................................70
7.RECOMMENDATIONSANDFUTUREWORK.............................................................71
8.REFERENCES.............................................................................................................72
Appendix1 ...................................................................................................................74
Appendix2 ...................................................................................................................76

10
ListofFigures

Figure1 ProposedmechanismofH
2
ScorrosiononFe...18
Figure2 LinearPolarizationResistanceCurve........27
Figure3 TheoreticalanodicpolarizationscanonStainlesssteel.......30
Figure4 Theoreticalcathodicpolarizationscan........31
Figure5 Tafelslopecalculation...31
Figure6 Nyquistplotwithonetimeconstantforthecircuitinfigure7.34
Figure7 Simplecircuitwithonetimeconstant......34
Figure8 Bodeplotwithonetimeconstant.......35
Figure9 Nyquistplotshowingthesolutionresistanceand
Polarizationresistance..35
Figure10 Bodeplotshowingsolutionresistanceand
Polarizationresistance..36
Figure11 TheGalvaniccell.........39
Figure12 DiagramoftheGalvaniccell..39
Figure13 Thechangeingalvaniccurrentwithtimeforvarious
concentrationofsulfideatpH3...42
Figure14Thegalvanicpotentialversustimeforvariousconcentration
ofsulfideatpH342
Figure15 Pictureofthecounterelectrodefortheexperiment
withaconcentrationofsulfide50mM.......43
Figure16 Pictureoftheworkingelectrodefortheexperiment
withaconcentrationofsulfide50mM....43
Figure17 Thechangeinpotentialatdifferentconcentration
ofsulfideatpH344
Figure18 Thepotentiodynamicsweepsforvariousconcentrationof
sulfideatpH3withbubblingN
2
..45
Figure19 EffectofconcentrationoncorrosionrateatpH3
measuredwithLPRandTafel..45

11
Figure20 Thegalvaniccurrentversustimeatvariousconcentrationof
sulfideinthepresenceofCO
2
..46
Figure21 Thegalvanicpotentialversustimeforvariousconcentrationof
sulfideatpH3inthepresenceofCO
2
..46
Figure22 Thechangeinpotentialatdifferentconcentrationof
sulfideatpH3inthepresenceofCO
2
..47
Figure23 Thepotentiodynamicsweepsforvariousconcentrationof
sulfide1mM,10mM,50mMatpH3withN
2
andCO
2
.47
Figure24 TheeffectofconcentrationoncorrosionrateatpH3
inthepresenceofCO
2
..48
Figure25 Thechangeingalvaniccurrentwithtimeforthe
concentrationofsulfide1mM,10mM,50mMatpH7.49
Figure26 Thegalvanicpotentialversustimeforvarious
concentrationofsulfideatpH7..........................................................49
Figure27 Thechangeinpotentialatdifferentconcentrationof
sulfideatpH7withbubblingN
2
.50
Figure28 Thepotentiodynamicsweepsforvariousconcentrationof
sulfide1mM,10mM,50mMatpH7.50
Figure29 Corrosionrateatvariousconcentrationofsulfide
atpH7measuredwithLPRandTafel...51
Figure30 Thegalvaniccurrentmeasuredfor20hoursatpH7with
concentrationofsulfideas1mM,10mM,50mM
inthepresenceofCO
2
....52
Figure31 Thegalvanicpotentialversustimeforvariousconcentrationof
sulfideatpH7inthepresenceofCO
2
52
Figure32 Thechangeinpotentialatdifferentconcentrationofsulfideat
pH7inthepresenceofCO
2
..53
Figure33 Thepotentiodynamicsweepsforvariousconcentration
ofsulfide1mM,10mM,50mMatpH7withN
2
andCO
2
53

12
Figure34 ThecorrosionratemeasuredwithLPRandTafel
atvariousconcentrationofsulfideforpH7inthe
presenceofCO
2
.............................................................................54
Figure35 Thegalvaniccurrentmeasuredfor20hoursforthe
concentrationofsulfide1mM,10mM,50mMatpH10..55
Figure36 Thegalvanicpotentialversustimeforvariousconcentration
ofsulfideatpH10...55
Figure37 Thechangeinpotentialatdifferentconcentrationof
sulfideatpH10.56
Figure38 Thepotentiodynamicsweepsforvariousconcentrationof
sulfide1mM,10mM,50mMatpH10withbubblingN
2
...56
Figure39 ThecorrosionratemeasuredwithLPRandTafelatpH10for
variousconcentrationofsulfide57
Figure40 Thegalvaniccurrentmeasuredfor20hoursinthepresence
ofCO
2
forvariousconcentrationofsulfide...58
Figure41 Thegalvanicpotentialversustimeforvariousconcentration
ofsulfideatpH10inthepresenceofCO
2
.....58
Figure42 ThechangeinpotentialatpH10forvariousconcentrationof
sulfideinthepresenceofCO
2
....59
Figure43 Thepotentiodynamicsweepsforvariousconcentrationof
sulfide1mM,10mM,50mMatpH10withN
2
andCO
2
..59
Figure44ThecorrosionratemeasuredwithLPRandTafelatpH10
inthepresenceofCO
2
....60
Figure45 TheeffectofpHongeneralcorrosionrate.....60
Figure46 ThepotentialpHdiagramforironinwaterat25C......61
Figure47 Theoreticalconditionsofcorrosion,immunityand
passivationofIron61
Figure48 CorrosionratemeasuredforblankwithLPRandTafel....62
Figure49 TheNyquistplotforCO
2
andH
2
Scorrosion..63
Figure50 TheNyquistplotforH
2
Scorrosion.....63

13
Figure51 TheBodeplotforCO
2
andH
2
Scorrosion64
Figure52 TheBodeplotforH
2
Scorrosion...64
Figure53 SummaryofcorrosionratemeasuredwithLPR.65
Figure54 SEMimageoftheelectrodeexposedtothesolutionpurged
withCO
2
.pictureAistakenatamagnificationof400XandpictureB
atamagnificationof2000X..67
Figure55 TheSEMpicturesoftheworkingelectrodewastakenfor
theexperiment50mMsulfideatpH10inthepresenceofCO
2
.
Thepictureshowsthefilmatvariousmagnification.a)200X,b)400X,
c)2030X,d)2000X...67
Figure56 SEMimageofthecrosssectionofthefilm....68
Figure57 TheSEMXrayanalysisofcrosssectionofthefilm.
ThepictureAistakennearthemetalsurface(bottomofthefilm)
andpictureBontopofthefilm....68
Figure58 TheSEMXrayanalysisofthesurfaceofthefilm.
Thisanalysiswasdoneatamagnificationof2000X.PictureAisfrom
a very rough surface of the film and Picture B is from a smooth
surfaceofthefilm69

14
ListofTables

Table1 TheExperimentaltestmatrix..37
Table2 TheChemicalcompositionofCarbonSteel..37
Table3 Summaryofcorrosionrate..65

15
1.INTRODUCTION
Corrosion of steel by CO2 and CO
2
/H
2
S has been one of the major problems in the
oil industry since 1940. Recently, it has again come to the fore because of the
technique of CO
2
injection for enhanced oil recovery and exploitation of deep
naturalgasreservoirscontainingcarbondioxide[1].

Thepresenceofcarbondioxide,
hydrogen sulphide (H
2
S) and free water can cause severe corrosion problems in oil
and gas pipelines. Internal corrosion in wells and pipelines is influenced by
temperature, CO
2
and H
2
S content, water chemistry, flow velocity, oil or water
wettingandcompositionandsurfaceconditionofthesteel.Asmallchangeinoneof
these parameters can change the corrosion rate considerably. In the presence of
CO
2,
the corrosion rate can be reduced substantially under conditions when
corrosion product, iron carbonate (FeCO
3
) can precipitate on the steel surface and
form a dense and protective corrosion product film. This occurs more easily at high
temperature or high pH in the water phase. When corrosion products are not
deposited on the steel surface, very high corrosion rates of several millimetres per
yearcanoccur.WhenH
2
SispresentinadditiontoCO
2
,ironsulphide(FeS)filmsare
formedratherthanFeCO
3
.Thisprotectivefilmcanbeformedatlowertemperature,
since FeS precipitates much easier than FeCO
3
. Localised corrosion with very high
corrosion rates can occur when the corrosion product film does not give sufficient
protection, and this is the most feared type of corrosion attack in oil and gas
pipelines.
Extensive studies had been done for CO
2
corrosion and H
2
S corrosion, but there is
very little understanding of the corrosion behaviour in the presence of both the
species. Hence, the objective of this project is to analyse the electrochemical
behaviourofcarbonsteelinthepresenceofbothCO
2
andH
2
S.
In order to fulfil this objective, classical electrochemical techniques like galvanic
effect, polarization techniques and electrochemical impedance spectroscopy are
used to find the corrosion rates in the CO
2
/H
2
S environment. The experiment is
performedatroomtemperatureandatdifferentpH.

16
2.LITERATUREREVIEW
2.1.CO
2
Corrosion
Carbon dioxide (CO
2
) corrosion is one the most studied form of corrosion in oil and
gasindustry.Thisisgenerallyduetothefactthatthecrudeoilandnaturalgasfrom
the oil reservoir / gas well usually contains some level of CO
2
. The major concern
withCO2corrosioninoilandgasindustryisthatCO
2
corrosioncancausefailureon
theequipmentespeciallythemaindownholetubingandtransmissionpipelinesand
thus can disrupt the oil/gas production. The basic CO
2
corrosion reaction
mechanismshavebeenwellunderstoodandacceptedbymanyresearchersthrough
theworkdoneoverthepastfewdecades.ThemajorchemicalreactionsincludeCO
2

dissolutionandhydrationtoformcarbonicacidasshowninequations(1)and(2),

) ( 2 ) ( 2 aq g
CO CO (1)

3 2 2 2
CO H O H CO + (2)

Thecarbonicacidthendissociatesintobicarbonateandcarbonateintwostepsasin
equations(3)and(4),

+
+
3 3 2
HCO H CO H (3)

+
+
2
3 3
CO H HCO (4)

CO
2
corrosion is an electrochemical reaction with the overall reaction given in
equation(5)

2 3 2 2
H FeCO O H CO Fe + + + (5)
Thus, CO
2
corrosion leads to the formation of a corrosion product, FeCO
3
, which
when precipitated could form a protective or a nonprotective scale depending on
theenvironmentalconditions[2].
The electrochemical reactions at the steel surface include the anodic dissolution of
ironasgiveninequation(6)

+
+ e Fe Fe 2
2
(6)

17
The cathodic reactions are proton reduction reaction and the direct reduction of
carbonicacidasinequations(7)and(8)

2
2 2 H e H +
+
(7)

+ +
3 2 3 2
2 2 2 HCO H e CO H (8)

Despite more than three decades of intense research, it is still not known which of
the two reactions (7) and (8) actually occur on the metal surface. Hence, the net
cathodic current was assumed to be the sum of the currents of the two cathodic
reactions. It has been suggested that the direct reduction of bicarbonate ion
becomesimportantathigherpH[3].
2.1.1TheeffectofpH
pHistheindicationoftheH
+
concentrationinthesolutions,whichisoneofthemain
speciesinvolvedinthecathodicreactionofCO
2
process.Ithasbeenillustratedboth
experimentally and computationally that corrosion rate changes significantly with
respecttopH.HigherpHleadstoadecreasedsolubilityofironcarbonateandthus
results in an increased precipitation rate, faster formation of protective films and
hencereductionofthecorrosionrate.
2.1.2Theeffectoftemperature
Temperature accelerates all processes involved in CO
2
corrosion including transport
of species, chemical reactions in the bulk of the solutions and electrochemical
reactionsatthemetalsurface.Thegrowthofironcarbonatefilmisaveryslowanda
temperature dependent process. Increasing the temperature increases the
precipitation rate of iron carbonate significantly. Depending on the solubility of
protective films, temperature can either increase or decrease the corrosion rate[4].
In the case of corrosion where protective films do not form (typically at low pH),
corrosion rate increases with increase in temperature. However, at a higher pH
increased temperature would accelerate the kinetics of precipitation and facilitate
protectivefilmformation,thusdecreasingthecorrosionrate.

18
2.2.H
2
SCorrosion
Theinternalcorrosionofcarbonsteelinthepresenceofhydrogensulfiderepresents
a significant problem for both oil refineries and natural gas treatment facilities.
Surfacescaleformationisoneoftheimportantfactorsgoverningthecorrosionrate.
Thescalegrowthdependsprimarilyonthekineticsofscaleformation.Incontrastto
relativelystraightforwardironcarbonateprecipitationinpureCO
2
corrosion,inan
H
2
S environment many types of iron sulfide may form such as amorphous ferrous
sulfide, mackinawite, cubic ferrous sulfide, smythite, greigte, pyrrhotite, troilite and
pyrite,amongwhichmackinawiteisconsideredtoformfirstonthesteelsurfacebya
direct surface reaction[5]. The poorly known mechanism of H
2
S corrosion makes it
difficulttoquantifythekineticsofironsulfidescaleformation.
A probable mechanism for Iron dissolution in aqueous solutions containing H
2
S
basedontheformationofmackinawitefilm,asproposedbySunetal[6]

isshownin
figure1.
adsorbed
S H Fe S H Fe
2 2
+ +

+
+ + +
adsorbed adsorbed adsorbed
H HS Fe S H Fe
2

Path1Path2

+ +
+ + +
ads ads ads ads
H FeHS H HS Fe
+ +
+ + + +
ads ads ads ads
H S Fe H HS Fe 2

+ +
+ + + e H FeHS H FeHS
ads ads ads ads

ads ads ads ads
H FeS H S Fe 2 2 + + +
+

ads ads ads ads

H FeS e H FeHS 2 + + +
+
Taylorspairing

Taylorspairing

e mackinawit ads
FeS S nFe nFeS
2 2
2
e mackinawit ads
FeS S nFe nFeS
2 2
2
Figure1.ProposedmechanismofH
2
ScorrosiononFe.

19
2.2.1.TheeffectofpH
The protective nature and composition of the corrosion product depend greatly on
thepHofthesolution.AtlowervaluesofpH(<2),ironisdissolvedandironsulfideis
not precipitated on the surface of the metal due to a very high solubility of iron
sulfidephasesatpHvalueslessthan2.Inthiscase,H
2
Sexhibitsonlytheaccelerating
effectonthedissolutionofiron.AtpHvaluesfrom3to5, inhibitiveeffectofH
2
Sis
seen due to the formation of ferrous sulfide (FeS) protective film on the electrode
surface[7].
2.2.2.TheeffectofH
2
SConcentration
H
2
S concentration has an immense influence on the protective ability of the sulfide
film formed. As the concentration of H
2
S increases, the film formed is rather loose
evenatpH35anddoesnotcontributetothecorrosioninhibitingeffect[8].
2.2.3.Theeffectoftemperature
ThetemperaturedependenceofH
2
Scorrosionisveryweakforshorttermexposure
anddoesnotseemstohaveaneffectatlongerexposuretimes.Thissuggestthatthe
corrosionrateispredominantlycontrolledbythepresenceofironsulfidescale[5].
2.3.CO
2
/H
2
SCorrosion
TheinternalcorrosionofmildsteelinthepresenceofbothCO
2
andH
2
Srepresentsa
significant problem for oil and gas industries. Although the interaction of H
2
S with
lowcarbonsteelshavebeenpublishedbyvariousauthors,theunderstandingofthe
effect of H
2
S on CO
2
corrosion is still limited because the nature of the interaction
withcarbonsteeliscomplicated.
In the presence of H
2
S, additional chemical reactions occurring in the bulk of the
solutioninclude:
DissociationofdissolvedH
2
Sisgiveninequation(9).

+
+ HS H S H
S H
K
2
2
(9)
where
| || |
| | S H
HS H
K
S H
2
2
+
=

20
ThedissociationofHS

ionisgivenbyequation10.

+
+
2
S H HS
HS
K
(10)
where
| || |

+
=

HS
S H
K
HS
2

H
2
S gas is about three times more soluble than CO
2
gas, the acid created by the
dissociation of H
2
S is about three times weaker than carbonic acid[9]. Hence, the
effectofH
2
SgasondecreasingthesolutionpHisapproximatelythesameasCO
2
gas.
Unlike dissolved CO
2
, dissolved H
2
S does not need to undergo the slow hydration
stepinordertobecomeanacid.
In a H
2
S dominated system, H
2
S lower the solution pH as it acts as a weak acid like
carbonicacid.Itcanalsoincreasethecorrosionrateinasimilarwayascarbonicacid,
byprovidinganextracathodicreactionasinequation(11)

+ + HS H e S H
2
(11)

However, this direct reduction of H

2
S is only feasible if the amount of H
2
S is high
enough, which means the system has to be either H
2
S dominated system (sour
regime) or CO
2
/H
2
S mixed system[10]. Moreover, elemental sulphur is often
associatedwithhighconcentrationofH
2
Sandverylittleisknownaboutthecomplex
interactionstakingplaceinthepresenceofelementalsulphur.
WhenhydrogensulfideispresentinlowconcentrationinaCO
2
dominatedsystem,
the iron sulfide (FeS) film interferes with the formation of the carbonate scale
(FeCO
3
)[11].Thisisofinterestbecausetheironsulfidefilmwouldseemtobemore
easily removed from the pipe wall than the iron carbonate scale. Under turbulent
conditions,removaloftheprotectivescalewillleadtoanincreasedcorrosionrate.
The kinetics of scale formation in the CO
2
/H
2
S system is complicated and still not
understood well. The makeup of the surface scale under these conditions will not
only depend on the chemistry of the brine and the respective solubility of iron
carbonates and iron sulfides, but also on the competitive kinetics of the two scale
formationmechanisms[12].

21
2.4.Corrosionproductfilmformation
CO
2
/H
2
S corrosion on the metal surface is strongly dependent on the type of
corrosion product film formed on the surface of the metal during the corrosion
process. The precipitation rate or the formation of these films depends on various
environmental factors and greatly on the concentration of species. The stability,
protectiveness, and adherence of these films determine the nature and the rate of
corrosion. Depending on the composition, the corrosion films can be of different
forms.
2.4.1.Ironcarbide(Fe
3
C)
Iron carbide is an undispersed component of mild steel, which is left behind after
the corrosion of iron from the steel structure. Iron carbide films are conductive
electrically, very porous and nonprotective[13] films can significantly affect the
corrosion process by either decreasing the corrosion rate by acting as a diffusion
barrier,orincreasingthecorrosionbyincreasingtheactivespecimensurfaceareaby
formingaconductivebridgebetweenthecounterandworkingelectrodes.Also,this
kind of film formation could result in galvanic coupling of the film to the metal or
acidificationofthesolutioninsidethecorrosionproductfilmwhichisverydangerous
and by far the strongest reason that could be given for the occurrence of localized
corrosion.
2.4.2.Ironcarbonate(FeCO
3
)
InaCO2corrosionsituation,ironcarbonateisformedfromthereactionofironand
carbonateionsgivenbyequation(12)

( ) s FeCO CO Fe
3
2
3
2
+
+
(12)

The rate of precipitation of iron carbonate is so slow that most often the
precipitation kinetics rather than the thermodynamics come into consideration.
Precipitationofsolidironcarbonateoccurswhentheproductoftheconcentrations
ofFe
2+
andCO
3
2
ionsinthesolutionexceedacertainlimitasthesolubilitylimit.

22
The rate of precipitation of the iron carbonate ( ) (
) ( 3 s
FeCO
R can be expressed by the
equation(13)[2]

) ( ) (
3 3 ) ( 3
FeCO FeCO FeCO
S f Ksp T f
V
A
R
S
=
(13)

where A/V is the surface area to volume ratio and Ksp

FeCO3
is the solubility limit of
FeCO
3.

ThesupersaturationSisdefinedasinequation(14)

3
2
3
2
3
FeCO
CO Fe
FeCO
Ksp
c c
S
+
=
(14)
SinceCO
3
2
ionconcentrationisdependentonthepH,itcanbededucedtoeqn.15
) , (
2
pH Fe f S
+
= (15)

Therefore, supersaturation and temperature are the most important factors

affecting the rate of precipitation, and the nature and protectiveness of the iron
carbonatefilm.Precipitationofironcarbonateonthesurfaceofthemetaldecreases
the corrosion rate by acting as a diffusion barrier for the corrosive species to travel
to the metal surface by blocking few areas on the steel surface and preventing
electrochemicalreactionsfromhappeningonthesurface[14].
2.4.3.Ironsulfide(FeS)film
The structure and composition of the protective FeS film depends greatly on the
concentrationofH
2
Sinthesystem.Theprotectivenatureofthefilmmainlydepends
on the pH of the solution [15]. At a solution pH value of 3 to 5, with a small
concentrationofH
2
S,aprotectivefilmofFeSinhibitsthecorrosionrateofthemetal
coupon[7]. In nearly neutral pH and at room temperature, mackinawite forms
throughasolidstatereaction,whileatapHvaluebetween5and7,amorphousFeS
precipitates. The kinetics of FeS formation is complicated than the iron carbonate
film.Thereactionfortheformationofsolidironsulphideisgiveninequation(16).
) (
2 2
s FeS S Fe +
+
(16)

23
Itisassumedthattheprecipitationofsolidironsulphideoccurswhentheproductof
the concentration of Fe
2+
and S
2
exceed the

solubility limit of FeS. The rate of

precipitationofIronsulfideisgiveninequation(17)

) ( ) (
) ( FeS FeS s FeS
S f Ksp T f
V
A
R =
(17)
WheresupersaturationS
FeS
isdefinedinequation(18)

FeS
S Fe
FeS
Ksp
c c
S
+
=
2 2
(18)
ItcouldalsobeobservedthatthesupersaturationofFeSisastrongfunctionofthe
concentrationofH
2
SandFe
2+
.Atthemetalsurface,becauseofthefastdepletionof
H
+
ions, the local pH near the metal surface is greater than the bulk pH. This could
resultinalargersupersaturationofFeSandprecipitationhappensevenfasteratthe
metal surface. Since iron sulfide is a semiconductive film, precipitation of iron
sulfide in combination with other kind of nonconductive film (e.g. FeCO
3
) on the
surfaceofthemetalcoulddecreasethecorrosionratebyactingasadiffusionbarrier
for the corrosive species to travel towards the metal surface. If FeS is the only one
material precipitating on the surface of the metal, even if the film is thick, the
corrosion rate may not be low because of the conductivity of the film. Also,
precipitation of only FeS on the surface could result in false depiction of corrosion
rate if electrochemical techniques are used. The reason is the interference of the
conductiveFeSfilmintheprocessofelectrontransfer.
The amount of Fe
2+
has an immense influence on the formation and nature of the
iron carbonate and iron sulfide scale[12]. Increased Fe
2+
concentration can lead to
higher super saturation of both iron carbonate and iron sulfide according to
equations (13) and (17), respectively, which could increase the precipitation rate of
iron carbonate and iron sulphide. The precipitated film could be very protective by
beingdenseandactingasadiffusionbarriertothecorrosivespecies,oritcouldbe
porous and thick and still could not be protective. Porous and incomplete films are
veryhazardoustothepipewallastheyareveryfavorableforlocalizedattackofthe
metal. Hence, porosity of the film is the most important factor in determining the
corrosionrateofthefilmunderfilmingconditions.

24
Researchers[12] have found that the corrosion products formed in CO
2
/H
2
S system
depends on the competitiveness of iron carbonate and mackinawite. At high H
2
S
concentration and low Fe
2+
concentration, mackinawite is the predominant scale
formed on the steel surface. At low H
2
S concentration and high Fe
2+
concentration,
bothironcarbonateandmackinawiteareformed.

25
3.ELECTROCHEMICALMETHODS
3.1GalvanicCorrosion
Galvanic corrosion, also referred to as twometal or bimetallic corrosion, occurs
when two dissimilar metals or alloys are in contact electrically while both are
immersedinanelectrolytesolution.Oneofthetwometalsiscorrodedpreferentially
by this type of corrosion; that is the most active or anodic metal corrodes rapidly
while the more noble or cathodic metal is not damaged. Galvanic attack can be
uniform in nature or localized at the junction between the alloys depending on
conditions. It can be particularly severe under the condition where protective
corrosion film does not form or where they are removed by condition of erosion
corrosion.
Every metal or alloy has a unique corrosion potential. E
corr
, when immersed in a
defined corrosive electrolytic solution. Thus, when two dissimilar metals are
connected in an aqueous environment, their differences in corrosion potentials will
causecorrosion.Themetalwiththemorenegativepotentialperformoxidationand
the other metal with more positive potential perform reduction. Thus, in a couple
betweentwometalsAandB,theactivemetalAistheanode,whilethenoblemetal
Bisthecathode,withthecorrespondingreactions:
AA
n+
+ne

(19)
B
m+
+me

B (20)

Everymetalhasbeenratedfornobilityandthenplacedongalvanicscalesaccording
tonobility.Basicallynobilityisanindicationoftheresistancetocorrosion,especially
of one metal contacting another metal. The relative nobility of a material can be
predicted by measuring its corrosion potential. The Galvanic series rank metals and
alloys in order of reactivity or electrical potential. Metals that are least noble are
very anodic, electropositive or high potential and will corrode most easily, whereas
metalsthataremorenoblearehighlycathodic,electronegativeorlowpotentialand
will be the more resistant to corrosion. The most corrosive effects will occur
between metals from the opposite ends of the galvanic scale or ranking of nobility.

26
Dissimilarmetalsincontactwitheachotherinthepresenceofanelectrolytecauses
currenttoflowthroughtheirpointsofcontactattheexpenseofthemetalwiththe
higherpotentialorlessnobility.Themuchlessnoblemetalisgraduallyconsumedin
the electrochemical reaction and will deteriorate or wear away as the metal ions
migrateawayfromtheveryanodicmetaltothemorenoblecathodicone.Themore
noble metal's corrosion resistance actually increases from this transfer of ions to it
fromthelessnoblemetal,whiletheothermetalisgraduallygettingconsumed.Also,
oxides formed on a metal surface can form a galvanic couple with the same metal
withnooxidefilmasthesetwometalsurfacecanhavedifferentpotential[16].
A zero resistance ammeter (ZRA) is used to measure the galvanic coupling current
between two dissimilar electrodes. ZRA is a current to voltage converter that
producesavoltageoutputproportionaltothecurrentflowingbetweenitstwoinput
terminalswhileimposingazerovoltagedroptotheexternalcircuit.
3.2LinearPolarizationresistance
The Linear polarization resistance method, based on electrochemical concepts,
enables determination of instantaneous interfacial reaction rates such as corrosion
ratesandexchangecurrentdensitiesfromasingleexperiment.
Wheneverthepotentialofanelectrodeisforcedawayfromitsvalueatopencircuit,
thatisreferredtoaspolarizingtheelectrode.Whenanelectrodeispolarized,itcan
causecurrenttoflowthroughelectrochemicalreactionsthatoccurattheelectrode
surface.Theamountofcurrentiscontrolledbythekineticsofthereactionsandthe
diffusionofreactantsbothtowardsandawayfromtheelectrode.
In cells where an electrode undergoes uniform corrosion at open circuit, the open
circuit potential is controlled by the equilibrium between two different
electrochemical reactions. One of the reactions generates cathodic current and the
otheranodiccurrent.

27

Figure2.LinearPolarizationResistanceCurve.

The open circuit potential ends up at the potential where the cathodic and the
anodic currents are equal. It is referred to as a mixed potential. The value of the
currentforeitherofthereactionsisknownasthecorrosioncurrent.Whenthereare
two simple, kinetically controlled reactions occurring, the potential of the cell is
relatedtothecurrentbyequation(21)

) (
) ( 303 . 2 ) ( 303 . 2
c
Eoc E
a
Eoc E
corr
e e I I
| |

= (21)
where,
Imeasuredcellcurrentinamps,
I
corr
corrosioncurrentinamps,
E
oc
opencircuitpotentialinvolts,

a
anodicBetacoefficientinvolts/decade

c
cathodicBetacoefficientinvolts/decade.

If a small signal is applied in approximation to equation (21), equation (22) can be

obtained
)
1
(
) ( 303 . 2
p c a
c a
corr
R
I
+
=
| |
| |
(22)

28
Where,R
p
polarizationresistance
IftheTafelconstantsareknown,I
corr
canbecalculatedfromR
p
usingequation(22).
I
corr
inturncanbeusedtocalculatethecorrosionrate.
3.2.1Calculationofcorrosionratefromcorrosioncurrent
ThecorrosioncurrentcanbeconvertedintocorrosionratebyusingFaradayslaw
nFM Q = (23)
Where
Qchargeincoulombs
nnumberofelectronstransferredpermoleculeoratom
FFaradaysconstant=96487.7coulombs/mole
Mnumberofmoles.
Equation (23) can be expressed in terms of equivalent weight (EW) by using the
relationsEW=Atomicweight(AW)/nandM=W/AW.TheexpressionforW,whichis
themassoftheelectroactivespecies,isgiveninequation(24)

F
Q EW
W

= (24)

Modifyingequation(24)givesequation(25)

A d
EW K I
CR
corr

= (25)

CRcorrosionrate.ItsunitsategivenbythechoiceofK
I
corr
corrosioncurrentinamperes
Kconstant=3272mm/yr
EWequivalentweightingrams/equivalent
Ddensityingrams/cm
3
Asampleareaincm
2
This formula is valid only for uniform corrosion. In cases where localized corrosion
occurs,thiscannotbeusedasitgivesverylowcorrosionratethanactuallyis.

29
3.3Potentiodynamicscan
Potentiodynamic polarization is a technique where the potential of the electrode is
variedataselectedratebyapplicationofacurrentthroughtheelectrolyte.Through
the DC polarization technique, information on the corrosion rate, pitting
susceptibility, passivity, as well as the cathodic behavior of an electrochemical
systemmaybeobtained.
Inapotentiodynamicexperiment,thedrivingforce(i.e.,thepotential)foranodicor
cathodicreactionsiscontrolled,andthenetchangeinthereactionrate(i.e.,current)
is observed. The potentiostat measures the current which must be applied to the
system in order to achieve the desired increase in driving force, known as the
applied current. As a result, at the open circuit potential the measured or applied
currentwillbezero.
3.3.1TheAnodicscan
Aschematicanodicpolarizationcurve,typicalforstainlesssteelisillustratedinfigure
2. As shown in figure 2, the scan starts from point 1 and progresses in the positive
(potential)directionuntilterminationatpoint2.Theopencircuitpotentialislocated
atpointA.Atthispotentialthesumoftheanodicandcathodicreactionratesonthe
electrodesurfaceiszero.TheregionBistheactiveregion,wheremetaloxidationis
the dominant reaction taking place. Point C is known as the passivation potential,
andastheappliedpotentialincreasesabovethisvaluethecurrentdensityisseento
decrease with increasing potential (region D), until a low, passive current density is
achieved(passiveregionregionE).Oncethepotentialreachedasufficientlypositive
value (point F, sometimes termed as breakaway potential) the applied current
rapidly increases (region G). This increase may be due to a number of phenomena,
depending on the alloy/environment combination. For some systems (e.g.,
aluminum alloys in salt water) thissudden increase in current may bepitting, while
forothersitmaybetranspassivedissolution.Forsomealloys,typicallythosewitha
veryprotectiveoxide,suchascobalt,thesuddenincreaseincurrentisduetooxygen
evolution.

30

Figure3.TheoreticalanodicpolarizationscanonStainlesssteel.

3.3.2CathodicScan
A schematic cathodic polarization scan is illustrated in figure 4. In a cathodic
potentiodynamic scan, the potential is varied from point 1 in the negative direction
topoint2.TheopencircuitpotentialislocatedatpointA.DependingonthepHand
dissolved oxygen concentration in the solution, region B may represent the oxygen
reduction reaction. Since this reaction is limited by how fast oxygen may diffuse in
solution (mass transport controlled) there will be an upper limit on the rate of this
reaction,knownaslimitingcurrentdensity.Furtherdecreaseintheappliedpotential
result in no change in the reaction rate, and hence the measured current remains
the same (region C). Eventually, the applied potential becomes sufficiently negative
foranothercathodicreactiontobecomeoperative,suchasillustratedatpointD.As
thepotential,andhencedrivingforcebecomesincreasinglylarge,thisreactionmay
become dominant, as illustrated in region E. This additional reaction is typically the
reduction of other species in the environment (such as the hydrogen evolution
reaction,alsoknownasthewaterreductionreaction).

31

Figure4.Theoreticalcathodicpolarizationscan.

3.3.3CorrosionratefromPotentiodynamicscan
For reactions which are essentially activation controlled, the current density can be
expressedasafunctionoftheoverpotential,,whichisexpressedinequation(26)

0
log
i
i
| q = (26)
Equation (26) is known as the Tafel equation, where is the Tafel slope, i is the
appliedcurrentdensity,andi
0
istheexchangecurrentdensity.

Figure5.Tafelslopecalculation.
Thus,theTafelslopefortheanodicandcathodicreactionsoccurringatopencircuit
may be obtained from the linear regions of the polarization curve, as illustrated in

32
figure 5. Once these slopes are established, it is possible to extrapolate back from
both the anodic and cathodic regions to the point where the anodic and cathodic
reactionrates(i.e.,currents)areequivalent.Thecurrentdensityatthatpointisthe
corrosion current density (i
corr
) and the potential at which it falls is the corrosion
potential (E
Corr
). The corrosion current density can then be used to calculate the
corrosionrateusingequation(25).

3.4ElectrochemicalImpedanceSpectroscopy
Alternating Current (AC) impedance or Electrochemical Impedance Spectroscopy
(EIS) technique is one of the most powerful techniques for defining reaction
mechanisms, for investigating corrosion process and for exploring distributed
impedance system. Most generally, the application of the EIS technique has been
used by researchers for the evaluation of corrosion inhibitors, anodic coatings and
polymeric coatings. A brief introduction to the measurement technique is given
below:
The ability of a circuit element to resist the flow of electrical current is called
resistance. The resistance of an ideal resistor is defined by ohms law as the ratio
betweenthevoltageEandcurrentIasinequation(27)

I
E
R = (27)
AnidealresistorfollowsOhmslawatallvoltageandcurrentlevelsanditsresistance
value is independent of frequency. Circuit elements which exhibit much more
complexbehaviorareencounteredinrealworldsituationswherethesimpleconcept
ofidealresistorcannotbeapplicable.Impedanceisamoregeneralcircuitparameter
which is similar to resistance in a way that it is also a measure of the ability of the
circuit to resist the flow of electrical current but it is more complicated in its
behavior.
Electrochemical impedance is usually measured by applying an AC potential to an
electrochemical cell and measuring the current through the cell. When we apply a

33
sinusoidalpotentialexcitation,theresponsetothispotentialisanACcurrentsignal.
Thiscurrentsignalcanbeanalyzedasasumofsinusoidalfunctions.
Electrochemicalimpedanceisnormallymeasuredusingasmallexcitationsignal.This
is done so that the cells response is pseudolinear. In a linear (or pseudolinear)
system,thecurrentresponsetoasinusoidalpotentialwillbeasinusoidatthesame
frequencybutshiftedinphase.
Theexcitationsignal,expressedasafunctionoftime,hastheformasequation(28)
) sin(
0
t E E
t
e = (28)
E
t
is the potential at time t, E
0
is the amplitude of the signal and is the radial
frequency. The relationship between radial frequency (expressed in
radians/second)andfrequencyf(expressedinhertz)isgivenbyequation(29)
rf t e 2 = (29)
In a linear system, the response signal, I
t
is shifted in phase () and has a different
amplitude,I
0
asgiveninequation(30)
) sin(
0
| e + = t I I
t
(30)
An expression analogous to Ohms law can be used to calculate the impedance of
thesystemasinequation(31)

) sin(
) sin(
) sin(
) sin(
0
0
0
| e
e
| e
e
+
=
+
= =
t
t
Z
t I
t E
I
E
Z
t
t
(31)
UsingEulersrelationshipinequation(32)
| | | sin cos ) exp( j j + = (32)
Theimpedanceisthenrepresentedasacomplexnumberasinequation(33)
) sin (cos ) (exp( ) (
0 0
| | | e j Z j Z
I
E
Z + = = = (33)
Where
I
E
Z =
0

Theexpressionforimpedanceziscomposedofbothrealandimaginaryparts.Aplot
ofrealpartofimpedanceonXaxisandnegativeofimaginarypartofimpedanceon
YaxisiscalledNyquistplot.Figure6showstheshapeofNyquistplotforthesimple
equivalent circuit with one time constant, as shown in figure 7. The impedance on
the Nyquist plot can be represented as a vector of length |Z|. The angle between
this vector and the Xaxis is called the phase angle . The major short coming of a

34
Nyquist plot is that the frequency used to create a particular data point cannot be
recognized. The semicircle shown in figure 6 is characteristic of a single time
constant (for example, a combination of an ideal capacitance with a single
resistance). EIS plots for real cases contain more than one time constant and often
onlyportionofoneormoreofthesemicirclesisseen.

Figure6.Nyquistplotwithonetimeconstantforthecircuitshowninfigure7.

Figure7.Simplecircuitwithonetimeconstant.
Another way of expressing the impedance is the Bode plot. In Bode plot the
impedanceisplottedwithlogfrequencyonthexaxisandboththeabsolutevalueof
theimpedance(|Z|=Z
0
)andphaseshiftontheyaxis.TheBodeplotfortheelectric
circuit of Figure 7 is shown in Figure 8. Unlike the Nyquist plot, the Bode plot
explicitlyshowsfrequencyinformation.

35

Figure8.Bodeplotwithonetimeconstant.
3.4.1Corrosionratefromimpedanceplot
In a Nyquist plot as shown in figure 9, at very high frequency, the imaginary
component, Z'' disappears, leaving only the solution resistance, R
s
. At very low
frequency, Z'' again disappears, leaving a sum of R
s
and the Faradaic reaction
resistance or polarisation resistance, R
p
. The corrosion rate can be calculated by
using the SternGeary equation by assuming a reasonable value for the beta
coefficients.
Low frequency
High frequency
Rs Rs+Rp
10 260 510 760 1010
0
100
200
300
400
500
600
700
Z' (Ohm)
Z
'
'

(
O
h
m
)

Figure9.NyquistplotshowingthesolutionresistanceandPolarizationresistance.

36
In a bode plot the solution resistance and the polarization resistance can be read
fromthemagnitudeplotasshowninfigure10andthencancalculatethecorrosion
rate.

Figure10.BodeplotshowingsolutionresistanceandPolarizationresistance.

The various electrochemical methods described above is been referred from

standardtextbooksandresearcharticlesbyvariousauthors[1719].

37
4.EXPERIMENTALPROCEDUREANDSETUP
Researchobjectives
Theobjectiveofthisprojectistostudythecorrosionbehaviorofcarbonsteelinthe
presenceofbothCO
2
andH
2
SindifferentpHandconcentration.Thetestmatrixfor
theresearchisgiveninTable1

Table1.TheExperimentaltestmatrix
Steeltype St523
Standardelectrolyte 0.5%NaCl
Temperature 22C(roomtemperature)
pH 310
Concentrationofsulfide 1mM50mM
Pressure 1bar

Carbonsteelisusedforthispurposebecauseitisoneofthemostwidelyusedmetal
in the oil and gas industry. Table 2 shows the chemical composition of carbon steel
(St523),whichwasusedfortheresearch.

Table2.ChemicalcompositionofCarbonSteel
Element Weight%
C 0.15
Si 0.30
Mn 1.20
P 0.019
S 0.01
Nb 0.002
Fe 98.319

The experiment was done in a galvanic setup with two carbon steel electrodes in
differentelectrolyticsolution.

38
The sample was prepared by cutting a carbon steel rod having a surface area of
0.785cm
2
intoanapproximatelengthof1cm.Awirewassolderedtothesampleand
it was molded with epoxy resin. The reference electrode used was a Ag/AgCl
(Ref201, Radiometer Analytical, France), which has a potential of 0.197V vs. the
standardhydrogenelectrode(SHE)[20].Theaccuracyofthereferenceelectrodewas
checkedeverytimebeforetheexperiment,againstthestandardAg/AgClelectrode,
thedifferencewas3mVorlessinallcases.
For the galvanic coupling two glass cells and a lid for each cell is made, and
appropriate holes are drilled in the lid for the electrodes, for a bridge and for
nitrogen purging. The cells are made air tight by applying grease between the cell
andthelid.Astheholesfortheelectrodeandthebridgearemarginallylargerthan
needed,aTeflontapeisusedtosealtheholesairtight.
Ofthetwocells,onecellisaddedwith0.5MNaClsolutionanditismaintainedatpH
7 while on the other cell the condition is varied. The two cells are connected by a
tube (bridge) and the solution is pumped into the tube to facilitate the flow of ions
betweenthecells.Tominimizecontaminationoftheblankcellwithsulfide,eachend
ofthebridgeisfilledwithcotton,sothatthesulfidegetsfilteredoffduringtheionic
transfer. Also, a thread is drawn all the way through the bridge in order to prevent
the formation of air bubbles and assure good electrolytic contact. Both cells are
purged with nitrogen to prevent oxidation. The galvanic corrosion effect is
monitored for 20 hours for each experiment. After 20 hours the galvanic setup is
disconnected and the LPR measurement is taken for the working electrode for 1
hour with platinum as the counter electrode. The potential range for the LPR
measurements are from 0.02V to 0.02V. The potentiodynamic polarization scans is
madewithaninitialvoltageof0.9Vandafinalvoltageof0.5Vwithascanrateof
0.15mV/s.Forgalvaniccorrosionmeasurementsthereferenceelectrodeisplacedin
the blank cell to prevent contamination of the electrode by sulfide. For LPR
measurementsandpotentiodynamicscanthereferenceelectrodewasintroducedin
thecellwiththeworkingelectrodeinasyringe.Thiswastoimprovethecontact,as
the potentiodynamic scan gave poor results when the reference electrode was
placed in the blank cell. The contamination in this case is minimized by introducing

39
some cotton in the tube connected to the syringe. Figure 11 and 12 shows the
experimentalsetupindetail.
AsthecarbonatefilmformedunderCO
2
environmentneedlongertimetoformthan
thesulfidefilminH
2
Senvironment,thesamplesarestoredinacontainerwithsaline
waterpurgedwithCO
2
.Thesamplesarekeptinthisenvironmentfor40daysbefore
thestartofthefirstexperiment.Thetemperatureoftheenvironmentwas20Cfor
the first 20 days and the temperature was raised to 40C for the next 20 days. The
raise in temperature is to increase the film formation rate, as higher temperature
enhancestherateofprecipitation.

Figure11.TheGalvaniccell.

Figure12.DiagramoftheGalvaniccell1)0.5MNaClsolution,2)Counterelectrode,
3)Referenceelectrode,4)Bridge,5)Workingelectrode,6)Platinumelectrode(for
LPRmeasurementandTafelscans),7)Theexperimentalsolution,8)Syringe(toinsert
referenceelectrodeduringLPRandTafelscans),9)Rubberbulb.

40
The experiments were performed with sodium sulfide as the source for H
2
S gas, as
the use of H
2
S gas directly needs elaborate safety measures. Hence the amount of
H
2
SproduceddependsonthepHofthesolution.Alltheexperimentsaredonewith
the Gamry Potentiostat. For galvanic measurement it was connected in zero
resistance ammeter (ZRA) mode, which means metal 1 is connected as working
electrode,metal2asthecounterelectrodeandthereferencetoreferenceelectrode
TheelectrodesarepolishedwithP120siliconcarbidepaperbeforethestartofeach
experiment. After each experiment an enlarged image of the working electrode is
taken. The SEM imaging was also done to study the surface characteristics of the
film.TheSEM(ScanningElectronMicroscope)EDX(EnergydispersiveXray)analysis
was done to find the elemental composition in the metal film. The accelerating
voltage applied for SEM imaging was 10kV, in order to get a clear surface structure
without damaging the surface film, as higher accelerating voltage gives high
resolutionbutunclearsurfacestructureandcandamagethefilm.
The electrochemical impedance spectroscopy (EIS) analysis was done for the
concentrationof10mMsulfideatpH7forafrequencyrangeof20,000to0.05Hz.

41
5.RESULTSANDDISCUSSION
The experimental results obtained are presented below based on the sulfide
concentrationandpH.
ExperimentalSeries1:
a. H
2
SCorrosionpH=3,ConcentrationofSulfide=1mM,10mM,50mM
b. CO
2
/H
2
SCorrosionpH=3,ConcentrationofSulfide=1mM,10mM,50mM
ExperimentalSeries2:
a. H
2
SCorrosionpH=7,ConcentrationofSulfide=1mM,10mM,50mM
b. CO
2
/H
2
SCorrosionpH=7,ConcentrationofSulfide=1mM,10mM,50mM
ExperimentalSeries3:
a. H
2
ScorrosionpH=10,ConcentrationofSulfide=1mM,10mM,50mM
b. CO
2
/H
2
ScorrosionpH=10,ConcentrationofSulfide=1mM,10mM,50mM
ExperimentalSeries4:
a. H
2
S and CO
2
/H
2
S corrosion EIS Analysis, Concentration of Sulfide= 10mM,
pH=7
Experimentalseries1a
The galvanic currents in experiments with various concentration of sulfide at pH 3
are shown in figure 13. The general corrosion rate was expectedto behigh at pH 3
and the presence of H
2
S is supposed to further increase the rate of corrosion. But,
the results obtained in the galvanic coupling shows negative current which means
that the working electrode (H
2
S environment) is more electronegative than the
counterelectrode.Thiscouldbebecauseofthefilmformationandhencethemetal
gettingpassive.Itwasfoundthattheopencircuitpotential(OCP)forthetwosurface
wasdifferent,ahigherOCPattheworkingelectrode(H
2
Senvironment)andalower
OCPatthecounterelectrode(blank).Infigure13,itcanbeseenthatfor30minutes
to 1 hour the current was positive, suggesting that the passivation is due to film
formation.Figure14showsthegalvanicpotential,whichisamixedpotentialofthe
workingelectrodeandcounterelectrode,measuredfor20hours.Itcanbeseenthat
the potential gradually decreases to the more negative region, driving the anodic
currenttowardsthecounterelectrode.

42
Work done by Han J et al [21] with iron carbonate film shows that in a galvanic
setup,underfilmformingconditionthebaremetalcorrodesoneormoreordersof
magnitude faster than the film protected area. The result obtained was in
agreement with this research, showing a high level of protection of the working
electrodeinagalvaniccouplingbut whenthecouplingisremovedthemetalshows
highdegreeofcorrosion.

Figure13.Thechangeingalvaniccurrentwithtimefor
concentrationofsulfideatpH3.

Figure14.Thegalvanicpotentialversustimeforvarious
concentrationofsulfideatpH3.

43
Figure 15 shows the picture of the counter electrode taken after the end of the
experimentwith50mMsulfideconcentration.Itshowsthatauniformcorrosionhas
occurred in the electrode. Figure 16 is the picture of the working electrode for the
same experiment. It can be seen that the surface is covered with a thick black film.
The corrosion rate of the two metal surfaces during the galvanic coupling was not
measured, but the result suggests that the film covered surface is protected by the
baremetalsurface.

Figure15.Pictureofthecounterelectrodefortheexperiment
withaconcentrationofsulfide50mM.

Figure16.Pictureoftheworkingelectrodefortheexperiment
withaconcentrationofsulfide50mM.

44
Figure 17 shows the potential measured with the various electrochemical
techniques. Ec(Rp) and Ec(Tafel) are measured only for the working electrode after
disconnectingthegalvanicsetup.AtpH3asshowninfigure17thegalvanicpotential
waslowerthanthepotentialmeasuredwithLPRandTafel.Whencomparedwiththe
potentialpH diagram (the Pourbaix diagram) for iron in figure 46 and 47, this
potentialisintheactiveregionofcorrosion.

Figure17.ThechangeinpotentialatdifferentconcentrationofsulfideatpH3.

Figure18showsthepotentiodynamicsweepsforvariousconcentrationofsulfide.It
canbeseenthatatahigherconcentrationofsulfide(50mM)thecorrosioncurrentis
increasing suggesting that, higher the concentration of sulfide more will be rate of
corrosion. Figure 19 shows the corrosion rate calculated with LPR and Tafel. In
general, the corrosion rate, calculated with LPR shows higher value than the rate
measuredwithTafel.But,thetrendisthesameforbothtypesofmeasurement.

45

Figure18.Thepotentiodynamicsweepsforvariousconcentration
ofsulfide1mM,10mM,50mMatpH3withbubblingN
2.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1mM 10mM 50mM
Concentrati onofsul fi de
c
o
r
r
.
r
a
t
e

(
m
m
/
y
r
)
LPR
Tafel

Figure19.EffectofconcentrationoncorrosionrateatpH3measured
withLPRandTafel.

Experimentalseries1b
Inthisseries,theexperimentsaredonewiththeelectrodeprecorrodedwithCO
2
for
various concentration of sulfide at pH 3. The result obtained in this series is also
similartoexperimentalseries1a.Animportantobservationhereisthatthecurrent
(figure20)wasnegativefromthebeginningoftheexperimentforallconcentration
of sulfides. This is because the electrode has an initial carbonate film when it was
exposed to sulfide, so the electrode gets immediate protection and the bare metal
starts corroding from the beginning. The potential (figure 21) measured with the

46
galvanic setup shows a gradual decrease in the potential, driving the corrosion
current in the opposite direction. Figure 22 shows the potential measured by all
three methods. It shows that the galvanic potential is lower than the potential
measured by LPR and Tafel. When compared with the Pourbaix diagram (figure 46
and 47) this potential without the galvanic coupling measured by LPR and Tafel are
wellintothecorrosionregion.

Figure20.ThegalvaniccurrentversustimeatpH3forvariousconcentrationof
sulfideinthepresenceofCO
2.

Figure21.Thegalvanicpotentialversustimeforvariousconcentrationofsulfideat
pH3inthepresenceofCO
2
.

47

Figure22.Thechangeinpotentialatdifferentconcentrationof
sulfideatpH3inthepresenceofCO
2
.

Figure23.Thepotentiodynamicsweepsforvariousconcentrationof
sulfide1mM,10mM,50mMatpH3withN
2
andCO
2
.

Figure 23 shows the potentiodynamic sweep of the working electrode at different

concentrationofsulfide.Itshowsagradualincreaseincorrosioncurrentforincrease
in the concentration of the sulfide. Figure 24 shows the corrosion rate calculated
withLPRandTafel.Therateofcorrosioninthepresenceofcarbonatefilmisfound
tobemore thantheonewithoutcarbonatefilm.Mostof theliterature[10] suggest
thatthecarbonatefilmformsaverystrongprotectivelayerandpreventscorrosion.
But,theresultofthisexperimentwasnotinagreementwiththistheory.

48
According to some recent research [2123], the passive carbonate film can be
depassivated by decrease in pH. Also, the carbon dioxide gas is not supplied
continuously into the experimental environment and this undersaturation could
have dissolved the iron carbonate film. As the film dissolves, the irregular surface
beneath the film provides more space for reaction and hence the corrosion rate is
veryhigh.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1mM 10mM 50mM
Concentrationofsulfide
c
o
r
r
.
r
a
t
e
(
m
m
/
y
r
)
LPR
Tafel

Figure24.TheeffectofconcentrationoncorrosionrateatpH3inthe
presenceofCO
2
.

Experimentalseries2a
InthisseriesofexperimenttheelectrochemicalmeasurementsaretakenatpH7for
various concentration of sulfide. In the galvanic coupling, both the metals are
immersedinsolutionwithpH7.Thedifferenceinenvironmentisjustbythesulfide
concentration. The galvanic measurement shows negative current (figure 25) which
meansthecounterelectrodeactsasanode.Foratleast1hourfromthestartofthe
experiment the current remains positive and then turn negative, suggesting the
formation of film. The galvanic potential measurements (figure 26) show that the
decreaseinpotentialisdrivingthecurrentintheoppositedirection.

49

Figure25.Thechangeingalvaniccurrentwithtimefortheconcentrationofsulfide
1mM,10mM,50mMatpH7.

Figure26.Thegalvanicpotentialversustimeforvariousconcentrationofsulfideat
pH7.

The potential measured for the working electrode without the galvanic coupling by
LPRandTafelshowssimilarresultsasgalvanicpotential.InthepHrangeof4to10,
the corrosion rate of iron is relatively independent of the pH of the environment
(figure 45). In this pH range the corrosion rate is governed largely by the rate at
which oxygen reacts with absorbed atomic hydrogen, thereby depolarizing the

50
surface and allowing the reduction reaction to continue. As the experiment was
done in presence of nitrogen the corrosion rate was very low for lower
concentration of sulfide. At neutral pH, sodium sulfide produces very little H
2
S,
depending on the concentration of the sulfide. Also Na
2
S form a metal precipitate
whichishighlyinsolubleandpreventsthemetalfromfurtherattack.Thus,itcanbe
seen that the rate of corrosion (figure 29) is fairly high for 50mM sulfide
concentration (1.5 mm/yr) and for lower concentration the corrosion rate is
negligible. The potentiodynamic scan (figure 28) also shows that higher
concentrationhashigherdegreeofcorrosion.

Figure27.ThechangeinpotentialatdifferentconcentrationofsulfideatpH7.

Figure28.Thepotentiodynamicsweepsforvariousconcentrationofsulfide1mM,
10mM,50mMatpH7withbubblingN
2
.

51
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1mM 10mM 50mM
Concentrationofsulfide
C
o
r
r
.
r
a
t
e

(
m
m
/
y
r
)
LPR
Tafel

Figure29.CorrosionrateatvariousconcentrationofsulfideatpH7measuredwith
LPRandTafel.

Experimentalseries2b
In this series of experiment, the electrochemical measurement for the electrode
coveredwithironcarbonatefilmisdoneforvariousconcentrationofsulfideatpH7.
The galvanic current (figure 30) was negative as in the previous experiments. The
galvanic potential (figure 31) decreases from the open circuit potential and
stabilizingatapotentialofaround0.710V,andgivescathodiccurrent.Thepotential
(figure 32) measured from LPR shows a higher potential than the galvanic,
suggestingthatthecorrosionrateislow.Asdiscussedinexperimentalseries2a,the
neutral pH and oxygen free environment does not have much effect on corrosion.
But,forhigherconcentrationofsulfideitshowssomecorrosioneffect(figure34).
The potentiodynamic scan shows different potential for different concentration of
sulfide (figure 33). This could be because of some minor changes in the
environmentalconditions.

52

Figure30.Thegalvaniccurrentmeasuredfor20hoursatpH7withconcentrationof
sulfideas1mM,10mM,50mMinthepresenceofCO
2
.

Figure31.Thegalvanicpotentialversustimeforvariousconcentrationofsulfideat
pH7inthepresenceofCO
2
.

53

Figure32.Thechangeinpotentialatdifferentconcentrationof
sulfideatpH7inthepresenceofCO
2
.

Figure33.Thepotentiodynamicsweepsforvariousconcentrationofsulfide1mM,
10mM,50mMatpH7withN
2
andCO
2
.

54
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1mM 10mM 50mM
Concentrationofsulfide
c
o
r
r
.
r
a
t
e

(
m
m
/
y
r
)
LPR
Tafel

Figure34.ThecorrosionratemeasuredwithLPRandTafelatvariousconcentration
ofsulfideforpH7inthepresenceofCO
2
.

Experimentalseries3a
InthisseriesofexperimenttheelectrochemicalmeasurementaretakenatpH10for
various concentration of sulfide. The galvanic corrosion current (figure 35) was
almost zero, which means there is very little corrosion on both the metal surface.
The OCP measured for the working electrode is lower than the counter electrode.
So,thecorrosioncurrentisdrivingtowardstheworkingelectrode,butatpH10this
decreaseinpotentialisintheregionofimmunity(figure46and47).Hence,boththe
surfaceremainsprotected.Anotherreasonforverylittlecorrosioncouldbebecause
of the use of sodium sulfide as the source of H
2
S gas. At higher pH sodium sulfide
maynotproduceH
2
SgasandthedominantspeciescouldbeS
2
,andtheabsenceof
H
+
ionsdecreasesthecorrosionrate.Alsothereactionproduct,formedathigherpH
is metal sulfides, which when precipitates on the surface of the metal gives
protection to the metal as these metal precipitates have extremely low solubility
[24]. It is also been observed by researches that the film formation rate increases
withpHwhilethecorrosionratedecreasesinelectrolytesequilibratedinahydrogen
sulfideenvironment[25].ThegeneralcorrosionrateatpH10isalsoverylow(figure
45).Evenwithoutthegalvaniccoupling,averylowcorrosionratewasobservedwith
LPR and polarization scan for all the concentration of sulfide measured. Hence, the
sulfide concentration does not have an effect in causing corrosion at this pH. The
potential(figure37)measuredwithLPRandTafelisverylow.Thismightbebecause

55
of the formation of the passive film formed which lowers the potential. When this
potentialiscomparedwiththePourbaixdiagram(figure46and47),itisneartothe
regionofimmunitywherenocorrosionoccurs.Infact,thecorrosionratemeasured
withTafelshowsalmostzeroforallconcentrationofsulfides.

Figure35.Thegalvaniccurrentmeasuredfor20hoursfortheconcentrationof
sulfide1mM,10mM,50mMatpH10.

Figure36.Thegalvanicpotentialversustimeforvarious
concentrationofsulfideatpH10.

56

Figure37.ThechangeinpotentialatdifferentconcentrationofsulfideatpH10.

Figure38.Thepotentiodynamicsweepsforvariousconcentrationofsulphide1mM,
10mM,50mMatpH10withbubblingN
2
.

57

Figure39.ThecorrosionratemeasuredwithLPRandTafelatpH10forvarious
concentrationofsulfide.

Experimentalseries3b
In this series of experiment the electrochemical measurements are taken for the
electrode precorroded with CO
2
at pH 10. The galvanic current (figure 40) here
shows positive, which means the working electrode (sulfide environment) is
corroding. As discussed in experimental series 3a, the sulfide and its reaction
product does not seem to increase the corrosion rate. But the results of galvanic
currentshowsmildcorrosioneffectontheworkingelectrode,itcouldbebecauseof
thedifferenceinareabetweenthetwoelectrodesandalsocouldbebecauseofthe
presence of carbonate film. The precorroded metal surface has more area
comparedtothebaremetalsurfacebecauseoftheirregularityofthesurface.Inthe
galvanic coupling the OCP measured for the working electrode was lower than the
counterelectrode.But,atpH10thispotentialwasintheregionofimmunity(figure
46and47).So,itcanbeassumedthatthecorrosioneffectintheworkingelectrode
couldbebecauseofthedifferenceinarea,whichismakingtheworkingelectrodeas
anode. The potential (figure 42) measured from all three methods were very low,
suggesting that the probability of corrosion at this pH is negligible (from Pourbaix
diagram figure 46 and 47). The corrosion rate (figure 44) measured with LPR shows
around 1.5mm/yr for the concentration of 10mM sulfide and 50mM sulfide. This
corrosionratemaynotbeduetotheconcentrationofsulfidebecause10mMsulfide
showsslightlyhigherrateofcorrosionthan50mMsulfide.Thishighercorrosionrate

58
than expected could be because of the irregular surface of the electrode. Also, the
reactionproductofsodiumsulfidemayhaveinterferedwiththecarbonatefilmand
might have caused a local condition near the metal surface, which enhances the
corrosionrate.

Figure40.Thegalvaniccurrentmeasuredfor20hoursinthepresenceofCO
2
for
variousconcentrationofsulfide.

Figure41.Thegalvanicpotentialversustimeforvariousconcentrationofsulfideat
pH10inthepresenceofCO
2
.

59

Figure42.ThechangeinpotentialatpH10forvariousconcentrationofsulfideinthe
presenceofCO
2
.

Figure43.Thepotentiodynamicsweepsforvariousconcentrationofsulfide1mM,
10mM,50mMatpH10withN
2
andCO
2
.

60
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1mM 10mM 50mM
Concentrationofsulfide
C
o
r
r
.
r
a
t
e

(
m
m
/
y
r
)
LPR
Tafel

Figure44.ThecorrosionratemeasuredwithLPRandTafelforvariousconcentration
ofsulfideatpH10inthepresenceofCO
2
.

Standardresults
As the experiments are done by varying the pH of the solution it is necessary to
analyzetheresultswithastandardone.Figure45showsthegeneralcorrosionrate
fordifferentpHandfigure46,47arethePourbaixdiagramforironinwaterat25C
[16].

Figure45.TheeffectofpHongeneralcorrosionrate.

61

Figure46.ThepotentialpHdiagramforIroninwaterat25C

Figure47.Theoreticalconditionsofcorrosion,immunityandpassivationofIron.

ResultsofBlank
A set of experiment was done for pH 3, 7 and 10 without the addition of sulfide to
compare the experimental results. Figure 48 shows the corrosion rate measured
with Tafel and LPR. The LPR and the Tafel measurements for the working electrode
aretakenafter20hoursofgalvaniccouplingwiththeneutralsolution.

62

0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
pH3 pH7 pH10
c
o
r
r
.
r
a
t
e
(
m
m
/
y
r
)
LPR
Tafel

Figure48.CorrosionratemeasuredforblankwithLPRandTafel.

ExperimentalSeries4
EISAnalysis
TheEISanalysiswasdoneaspartoflearningthistechniquetomeasurethecorrosion
rate. This technique was used to measure the corrosion rate for 10mM sulfide
concentration at pH 7 with and without the presence of CO
2
. The frequency range
used for this technique was 20,000Hz to 0.05Hz. The Nyquist plot for the
measurement with carbonate film is shown in figure 49. The measurement was
taken initially in the presence of only CO
2
and N
2
and then after adding sodium
sulfide. The lower frequency range is not small enough to determine exactly the
point where the line crosses the intercept. Hence, the line is extrapolated to
determine approximately the corrosion rate. The corrosion rate for the carbonate
film covered surface before adding sodium sulfide was about 3mm/yr (the
coefficients was assumed to be 0.12V/decade for the corrosion rate calculation),
which is very high at this pH. This suggest that the film formed has dissolved in the
given environment and the irregular surface underneath was corroding at a higher
rate. From figure 49 it can be seen that the corrosion rate tends to decrease after
theadditionofsodiumsulfide.Butasthedataobtainedwasinsufficienttocalculate
exactly the corrosion rate it can be assumed to be around 1mm/yr. In the previous
experiments at pH 7, the corrosion rate was not measured during the start of the
experiment,buttheratecalculatedafter20hrs(figure34)was0.8mm/yr.

63
0
10
20
30
40
50
60
70
8 18 28 38 48 58
Real(ohm)
-
I
m
a
g
(
o
h
m
)
CO2
After H2S

Figure49.TheNyquistplotforCO
2
andH
2
Scorrosion.

Figure50showstheNyquistplotfor10mMsulfideconcentrationinthepresenceof
nitrogen at pH7. Extrapolating the curve, the corrosion rate was calculated to be
0.86mm/yr, at the beginning of the experiment and this rate is reduced to about
0.54mm/yrafter20hours.Thissuggeststhattheformationoftheprotectivefilmis
reducing the corrosion rate. Figure 51 and 52 shows the Bode plot for the same
experiment.Asthereisverylittledataavailableinthelowfrequencyregionitisnot
possible to calculate the corrosion rate. Also, as the Bode plot cannot be
extrapolatedtogetareasonablyaccurateresult.
0
50
100
150
200
250
300
350
400
450
0 100 200 300 400 500 600
Real (ohm)
-
I
m
a
g

(
o
h
m
)
H2S
After 20 hrs

Figure50.TheNyquistplotforH
2
Scorrosion.

64
0
10
20
30
40
50
60
70
2.0000 0.0000 2.0000 4.0000 6.0000
Logfrequency
R
e
a
l
(
o
h
m
)
CO2
AfterH2S

Figure51.TheBodeplotforCO
2
andH
2
Scorrosion.

Figure52.TheBodeplotforH
2
Scorrosion.

65
1mMS
1mMS
1mMS
1mMS/CO2
1mMS/CO2
1mMS/CO2
10mMS
10mMS
10mMS
10mMS/CO2
10mMS/CO2
10mMS/CO2
50mMS
50mMS
50mMS
50mMS/CO2
50mMS/CO2
50mMS/CO2
0 1 2 3 4 5
3
7
10
Corrosionrate(mm/yr)

Figure53.SummaryofcorrosionratemeasuredwithLPR.

Table3.Summaryofcorrosionrate.
H
2
Ssystem CO
2
/H
2
Ssystem Blanksystem Sulfide
concentration
LPR
(mm/yr)
Tafel
(mm/yr)
LPR
(mm/yr)
Tafel
(mm/yr)
LPR
(mm/yr)
Tafel
(mm/yr)
1mM 0.711 0.294 0.685 0.045
10mM 1.293 0.454 2.743 0.736
pH3
50mM 3.683 1.846 4.699 2.933
0.650 0.450
1mM 0.091 0.023 0.152 0.029
10mM 0.096 0.023 0.762 0.091
pH7
50mM 1.498 0.292 0.762 0.222
0.170 0.090
1mM 0.050 0.002 0.787 0.073
10mM 0.040 0.005 1.575 0.292
pH10
50mM 0.271 0.018 1.498 0.182
0.058 0.005

Figure 53 shows the summary of corrosion rate measured with LPR at different pH
levels for various concentration of sulfides. Table 3 summarizes the corrosion rate
calculatedwithLPRandTafel.

66
Ingeneral,itwasfoundthatthecorrosionrateishighintheCO
2
/H
2
Ssystemthanit
iswithjustH
2
S.Thiscouldbebecauseofthefollowingreasons:
- ForCO
2
/H
2
Scorrosionthecarbondioxidegasisnotsuppliedcontinuouslyin
the experimental environment. The experiment was done on precorroded
electrodewithCO
2
.
- The passivation due to iron carbonate film was lost when the metal was
introduced in the H
2
S dominated system which has an undersaturation of
carbondioxide
- As the experiment was done on a precorroded electrode there is more
surfaceavailableforH
2
Scorrosionthantheelectrodewhichisnotpreviously
corrodedwithCO
2

- Theexperimentwasdoneforashortperiodoftime.Thelongtermexposure
couldhavegivendifferentresult.
- TheH
2
Sgasisnotaddeddirectly,itisaddedintheformofsodiumsulfide.

SEMAnalysis
TheSEManalysiswasdonetostudythesurfacecharacteristicsofthefilm.Figure54
showstheSEMimageofthecarbonatefilm.Athighermagnificationthefilmshows
lots of cracks, which was not visible with naked eye. This crack is most probably
caused due to the drying of the sample while imaging in the SEM. Figure 55 shows
the SEM image of the film formed in the presence of both CO
2
and H
2
S. The metal
chosen for this imaging is the working electrode of the experiment with 50mM
sulfide at pH 10. This electrode was chosen because the thickness of the film was
maximum at pH 10. The surface topography of figure 54 is different from that of
figure 55. The carbonate film (figure 54) is more rough or crystalline than the film
formed in the presence of sulfide. In figure 54, images (c) and (d) are of almost the
same magnification but the aperture size for (c) is 30 m and for (d) is 60m. The
aperture size is increased to see more clearly the surface characteristics, as smaller
aperturecangivehighdepthoffieldbuttheresolutionisnotsogood.

67

Figure54.SEMimageoftheelectrodeexposedtothesolutionpurgedwithCO
2
.
PictureAistakenatamagnificationof400XandpictureBata
magnificationof2000X

Figure55.TheSEMpicturesoftheworkingelectrodewastakenfortheexperiment
50mMsulfideatpH10inthepresenceofCO
2
.Thepictureaboveshowsthefilmat
variousmagnification.a)200X,b)400X,c)2030X,d)2000X.

68
Figure 56 shows the cross section of the film at a magnification of 500X. The film
formedwasthickandthethicknessofthefilmwasapproximately100m.Figure57
showstheEDXanalysisofthefilmdonenearertothesurfaceofthemetalandatthe
top of the film. It is found that at the surface of the film there was about 9% by
weight of sulfur and at the bottom of the film it was 7.7%. This suggests that the
carbonatefilmwasmoreporousandthesulfidehaspenetratedthroughthefilmand
attacked the metal surface. But, this result cannot be considered as accurate as
sulfurisalowatomicweightspecies.Thepresenceofsiliconintheanalysisisdueto
thepolishingofthesamplewithsiliconcarbidepaper

Figure56.SEMimageofthecrosssectionofthefilm.

Figure57.TheSEMXrayanalysisofcrosssectionofthefilm.ThepictureAistaken
nearthemetalsurface(bottomofthefilm)andpictureBontopofthefilm.

69

Figure 58 shows the surface analysis done on the surface of the film. This analysis
wasdoneatamagnificationof2000X.Itwasfoundthatattheroughsurfaceofthe
filmtheconcentrationofsulfideisabout15%byweightandatthesmoothsurfaceit
was about 6% by weight, suggesting that the roughness of the surface is caused
becauseofthesulfidefilm.

Figure58.TheSEMEDXanalysisofthesurfaceofthefilm.Thisanalysiswasdoneat
amagnificationof2000X.PictureAisfromaveryroughsurfaceofthefilm.PictureB
isfromasmoothsurfaceofthefilm.

TheEDXanalysisforthecompositionofelementsisgivenindetailinAppendix2.

70
6.CONCLUSION
Fromtheexperimentsperformedoncarbonsteel(St523)withgalvaniccoupling
andwithoutthecoupling,thefollowingconclusionscanbemade:
- In a galvanic coupling between metals, the metal exposed to even higher
concentration of sulfide and lower pH gets protected and the bare metal
whichisinneutralpHwascorroding.
- The corrosion rate in general was high for lower pH levels and higher
concentrationofsulfides.
- ThecorrosionrateishigherfortheCO
2
/H
2
Ssystemthaninthepresenceof
onlyH
2
S.
- The Iron carbonate film formed appears to be a loose film if the
concentration of the carbonate in the solution is not saturated. This
conditionismoreobservedatlowerpH.

71
7.RECOMMENDATIONSANDFUTUREWORK
As this thesis is a short term project, many aspects related to the topic had to be
overlooked because of the time limitation. The following could be some of the
recommendationsforfuturework
- FurtherstudyisneededonCO
2
andH
2
Scorrosionforhighertemperatureand
pressure, as the temperature and pressure are usually higher in oil and gas
pipelines.
- All the experiments were done for a very limited period of 2024 hours. A
longtermeffectofthespeciesoncarbonsteelcanbeinvestigated.
- Further study can be done by supplying CO
2
gas continuously into the
experimentalenvironment.

72
8.REFERENCES

1. Zhao G.X, et al., 'Formation Characteristic of CO2 Corrosion Product layer of

P110 Steel Investigated by SEM and Electrochemical Technique'. Iron and
steelresearchInternational,Year2009.16(4):p.8994.
2. Hunnik E.W.J, V., Pots.B.F.M, and Hendriksen.E.L.J.A, The Formation of
Protective FeCO
3
Corrosion Product Layers in CO
2
Corrosion. Corrosion/96,
1996(PaperNo.6).
3. GrayL.G.S,etal.,EffectofpHandTemperatureontheMechanismofCarbon
Steel Corrosion by Aqueous Carbon Dioxide. Corrosion/90, 1990. Paper
No.40(Houston,TX:NACEInternational,1990).
4. Nesic S. and Lee K.L.J, A Mechanistic Model of Iron Carbonate Film Growth
andtheEffectonCO
2
CorrosionofMildSteel.Corrosion/02,2002.PaperNo.
237.
5. SunWandNesicS,AMechanisticModelofH
2
SCorrosionofMildSteel.NACE
Corrosion,2007(PaperNo.07655).
6. SunW,MarquezA.I,andBotteG.G,TheoreticalInvestigationofH
2
SCorrosion
ofMildSteel.InstitiuteofCorrosionandMultiphaseTechnology.
7. Ma H.Y, et al., The influence of hydrogen sulfide on corrosion of iron under
differentconditions.CorrosionScience,1999(42):p.16691683.
8. Shoesmith D.W, et al., The Formation of Ferrous Monosulfide Polymorphs
during the Corrosion of Iron by Aqueous Hydrogen Sulfide at 21
0
C.
ElectrochemicalSociety,1980.127(5):p.10071015.
9. RendonR.LandAlejandreJ,MolecularDynamicsSimulationsoftheSolubility
ofH
2
SandCO
2
inWater.J.Mex.Chem.Soc2008.52(1):p.8892.
10. Hausler R.H and Godard H.P, Advances in CO
2
Corrosion. NACE Publication,
1984.1.
11. Srinivasan S and Kane R.D, Prediction of Corrosivity of CO
2
/H
2
S Production
Environment.Preventionofpipelinecorrosionconference,Houston,TX,1995.
12. SunWandNesicS,KineticsofIronSulfideandMixedIronSulfide/Carbonate
Scale Precipitation in CO
2
/H
2
S Corrosion. Corrosion NACExpo, 2006(Paper
No.06644).
13. Wang H.B, et al., Characterization of Inhibitor and Corrosion Product Film
using Electrochemical Impedance Spectroscopy(EIS). Corrosion 2001,
2001(PaperNo.01023).
14. Nordsveen M, et al., A Mechanistic Model for Carbon Dioxide Corrosion of
Mild Steel in the Presence of Protective Iron Carbonate FilmsPart 1: Theory
andVerification.CorrosionScience,2003.59:p.443456.
15. Ma H.Y, et al., Theoretical Interpretation on Impedance Spectra for Anodic
Iron Dissolution in Acidic Solutions Containing Hydrogen Sulfide. NACE
Corrosion1998.54(8).

73
16. Pourbaix M, Atlas of Electrochemical Equilibria in Aqueous Solutions. Nace
International.,1966.
17. Enos D.G and Scribner L.L, The Potentiodynamic Polarization Scan. Technical
Report33,1997.
18. Machiels A and Munson D, Predictive Model for Galvanic Corrosion. Electric
powerresearchinstitute,2004.
19. LasiaA,ElectrochemicalImpedanceSpectroscopyanditsApplications.Kluwer
AcademicPublishers,1999.32:p.143248.
20. Lundstrom M, Aromaa J, and Forsen O, Redox potential characteristics of
cupric chloride solutions. Science Direct Hydrometallurgy, 2008. 95: p. 285
289.
21. Han J, et al., Roles of Passivation and Galvanic Effects in Localized CO
2

CorrosionofMildSteel.Corrosion2008,NACEInternational,2008.PaperNo.
08332.
22. HanJ,etal.,ElectrochemicalInvestigationofLocalizedCO
2
CorrosiononMild
Steel.corrosion2007,NACEInternational,2007(paperNo.07323).
23. SunW,KineticsofIronCarbonateandIronSulfideScaleFormationinCO
2
/H
2
S
Corrosion.PhDdissertation,2006.
24. David R.B, Sodium Sulfides. KirkOthmer Encyclopedia of Chemical
Technology,1997.
25. Hemmingsen T, Hilbert L, and Nielsen L.V, Assessment of sulphur and H
2
S
corrosion by use of simultaneous ER, galvanic and optical measurements.
Eurocorr2003.

74
Appendix1

The result obtained for the galvanic coupling of metals was speculative, as the
expectedresultwashighercorrosioninthesulfideenvironmentthanforthemetalin
theneutralpHenvironment.Inordertoverifytheauthenticityoftheresultsandto
checkwhethertheZeroResistanceAmmeterisconnectedinthecorrectdirectionin
the Gamry potentiometer, a set of experiments was done for which the results are
known.

Galvaniccouplingbetweencarbonsteelandduplexsteel
Agalvaniccouplingwasmadebetweencarbonsteelandduplexsteelwiththesame
experimental setup and neutral pH. The carbon steel was made the working
electrode(metal1)andtheduplexsteelasthecounterelectrode(metal2).Figure1
shows the galvanic current measured for 20 hours. As expected the current was
positiveanditisfairlyhigh,suggestingthatthecarbonsteeliscorroding.

Figure1.Thegalvaniccurrentmeasuredforcarbonsteelversusduplexsteel.

75
Figure 2 shows the galvanic potential measured for 20 hours between carbon steel
andduplexsteel.Theopencircuitpotentialmeasuredforcarbonsteelislowerthan
the duplex steel and hence the carbon steel acts as anode and the duplex steel as
cathode.Thepotentialmeasuredisfairlyconstantfor20hours.

Figure2.Thegalvanicpotentialmeasuredforcarbonsteelvs.duplexsteel.

76
Appendix2

TheSEMEDXAnalysis
TheSEMEDXanalysisisdonetostudythesurfacecharacteristicsofthefilm.Figure
3and4giveaclearpictureoftheelementalcompositionofthecrosssectionofthe
film.

Figure3.TheSEMEDXanalysisofthecrosssectionofthefilm.Thisanalysisisdone
atthebottomofthefilmi.e.,nearthemetalsurface.

77

Figure4.TheSEMEDXanalysisofthecrosssectionofthefilm.Thisanalysisisdone
atthetopofthefilmsurface.

Figure5and6showstheSEMEDXanalysisofthesurfaceofthefilm.Figure5shows
the elemental composition at a rough surface of the film and figure 6 near the
smoothsurface.

78

Figure5.TheSEMEDXanalysisofthesurfaceofthefilm.Thisanalysisisdoneonthe
roughsurfaceofthefilm,whentheimagemagnificationis2000X.

79

Figure6.TheSEMEDXanalysisofthesurfaceofthefilm.Thisanalysiswasdoneona
smoothsurfaceofthefilmwhentheimagemagnificationwas2000X.