Int. J. Electrochem. Sci.

, 10 (2015) 10832 - 10846

International Journal of
ELECTROCHEMICAL
SCIENCE
www.electrochemsci.org

Investigations on Corrosion Behaviour of Welded Joint in

ASTM A355P5 Alloy Steel Pipe
Alin Constantin MURARIU1,*, Nicoleta PLEŞU2,*
1
ISIM Timisoara, 30 Mihai Viteazul Blv. 300222 Timisoara, Romania
2
Romanian Academy, Institute of Chemistry, Timisoara, 24 Mihai Viteazul Blv. 300223 Timisoara,
Romania
*
E-mail: [email protected], [email protected]

Received: 8 September 2015 / Accepted: 30 September 2015 / Published: 4 November 2015

In this paper an investigation on corrosion behaviour of welded joints was performed to facilitate
determination the causes of preferential degradation by corrosion of the welded metal (WM) and the
heat affected zone (HAZ), in a welded pipeline of grade ASTM A355P5, used in a petrochemical plant
as transfer line between a thermal cracking heater and a column which are part of an atmospheric
distillation and vacuum (DAV) unit. Mechanical testing, structural and chemical analyses were
performed and a heat treatment procedure was applied to reduce the corrosion of WM and HAZ.
Besides classical tests and analyses, electrochemical tests were performed in order to study the
influence of heat treatment on the welded joints. Measurements were carried out at the open-circuit
potential (OCP) on specimens immersed in 3% NaCl solution to observe the change of potential and
qualitatively appreciate the dissolution and/or creation of a protective layer on the component surface.
It was found that the corrosion of the weld analyzed in this research has resulted due to its dendritic
structure (pearlite, ferrite, complex carbide, bainite and martensite in localized areas), which reveals
the lack of post welding heat treatment (PWHT). A promising solution to this problem represents
proper PWHT. The corrosion resistance of HAZ and WM increases after temperature treatment, as a
result of the structure improvement as well as hardness reduction in the weld and adjacent areas.

Keywords: Corrosion, Pipeline, Petrochemical equipment, Structural Integrity, Failure

1. INTRODUCTION

Petrochemical equipment and systems are composed of complex components at which age,
wear, and generally the uncertainty linked to their technical condition are the main issues facing their
management system, considering new safety and environmental protection requirements.
Int. J. Electrochem. Sci., Vol. 10, 2015 10833

Iron and steel alloys are materials commonly used in manufacturing a series of industrial
equipments including pipelines because they offer excellent weldability, fine mechanical properties
etc.[1,2] The use of heat resistant chromium-molybdenum steels for manufacture of structures
operating at higher temperatures (up to +475 °C) is caused by their low cost compared to the top
performing steels. A335 is often called chrome-moly pipe because the chemical composition contains
Molybdenum (Mo) and Chromium (Cr). Molybdenum adds strength, elastic limit and resistance to
wear, impact properties, and hardness to steel, increases the resistance to softening, restrains grain
growth and makes chromium steel less susceptible to embrittlement. It also enhances the high
temperature creep strength, corrosion resistance and inhibits pitting of steel. As a result of the very
high costs associated to their replacement with completely new construction, remaining life assessment
and exploitation prolongation in safety condition of components, failure analysis and prevention [1],
especially referring to those working in severe stress conditions (high temperatures and pressures,
vibration, aggressive environment), are still the current topics of research in the present.
A key problem in the applications of these materials in the pipeline industry remains the
corrosion of these materials in the presence of aggressive factors [3 - 7], because of the cost and
technical difficulties in repairing the corroded parts [8]. Since, in some cases, indirect costs of
corrosion were more or less equal to direct costs [2], corrosion is the most significant cost related
technical challenge that provides strong motivation for researchers to understand and identify the
corrosion mechanisms, predict the corrosion behaviour and find appropriate solutions for preventing
and mitigation of corrosion effects on susceptible components. Corrosion of metals and alloys in
aqueous environments or other ionically conducting liquids is mostly electrochemical type. It takes
place when various electrochemical reactions occur on a metal surface. One of these reactions results
in the change of the metal or some elements in the metal alloy from a metallic state into a non-metallic
state. Corrosion products consist of solid or dissolved species. There are well-known different types of
corrosion, including “Localized Corrosion”, “Generalized Corrosion” and “Metallurgically Influenced
Corrosion”. The corrosion behaviour of metals and alloys and their associated mechanisms are
influenced by the chemical composition, metallographic structure, mechanical characteristics, stress
level and impurities.
Welding is a necessary and frequently used method for pipeline assembly. The microstructure
and the mechanical properties of materials subjected to a welding process are modified and usually
three evident zones can be highlighted [3, 4]. One is associated with the unchanged part of the welded
joint with properties as the base metal (BM). An intermediary zone is represented by the base metal
area affected by the generated heat due to the welding process, followed by a re-cooling process,
known as the heat affected zone (HAZ). In HAZ, the microstructure and properties can be critically
altered. The third zone is represented by the weld metal (WM). Microstructure of the WM is a specific
moulding complex structure. Thus, WM is a dendritic structure that solidified from a molten state.
Adjacent to the WM is the fusion line that makes the transition to HAZ. In multipass welding, the
microstructure of these zones could be affected by thermal field produced due to successive weld
passes. Thus, welding produces localized changes in the composition, structure, stress level and
impurities in the joined metals and alloys and implicitly can also affect corrosion resistance.
Int. J. Electrochem. Sci., Vol. 10, 2015 10834

The corrosion products can form a layer capable to protect further the steel surface of the
generated species through physically blocking the effect of corrosive attack [9, 10]. The corrosion of
different steel grades and in different aggressive environments has been presented in [11-14]. Chloride
is an excellent rust cause with a little moisture. The protective layer produced on the metal surface
frequently fails in the presence of the chloride ions in different pH [1, 13, 14]. In sodium chloride
solution the iron corrodes through dissolution into ferrous and ferric cations. Chloride ions after
adsorption on the base metal surface penetrate the protective layer mainly at its point defects [13].
Metallurgical effects could affect the corrosion resistance of carbon steel weldments [15], thus
preferential corrosion appear in HAZ or WM. Also, geometrical shape of the component, associated
with stress concentration factor, play an important role on corrosion resistance behaviour. Usually,
alloys with high hardness are most severely damaged by corrosion, while those that are overaged are
fairly less susceptible [16]. Thus, in case of welded structures, the post weld heat treatments (PWHT)
are beneficial since they control the hardness characteristics and the size and distribution of
intergranular precipitates. Due to the exploitation condition and pollutants that could be found in the
system, such as: hydrogen sulphide, gaseous hydrocarbons, carbon monoxide and carbon dioxide,
High Temperature Hydrogen Attack (HTHA) and Stress Corrosion Cracking (SCC) phenomena
[17,18] could affect the material of pipelines.
The goal of this paper is to study the corrosion nature of a pipeline which transports light
products by mechanical, optical and electrochemical methods. The results indicate the importance of
PWHT. An improper PWHT presents a direct negative impact on the operation safety of inspected
components by increasing the risk of failure.

2. EXPERIMENTAL

2.1 Material and methods

The experimental study was performed on samples taken from the transfer line Ø508 x 15.09
mm between the thermal cracking heater H1 and the column C1 of the DAV installation, which
generates light products (gasoline, diesel or gas) by atmospheric and vacuum distillation of crude oil.
A part of these pipeline welds were affected by preferential corrosion at the weld and HAZ. The
pipeline is working under a temperature of 355 °C and pressure of 0.3 MPa and is made of feritic
alloy-steel according to ASTM A355P5 - Standard Specification for Seamless Ferritic Alloy-Steel Pipe
for High-Temperature Service. Two coupons were taken from the transfer line, each containing both a
portion of pipe Ø508 x 15.09 mm and a portion of the welded joint and a portion of an elbow 90R =
1.5D, Ø508 x 15.09 mm, placed between the two welds, Figure 1.
Int. J. Electrochem. Sci., Vol. 10, 2015 10835

a) Coupon containing a portion of the weld no.23 b) Coupon containing a portion of the weld no.24
between elbow 90 R=1.5D, Ø508×15.09mm and between elbow 90 R=1.5D, Ø508×15.09mm and
pipe Ø508×15.09mm pipe Ø508×15.09mm

Figure 1. Samples for the experimental program, taken from desalinate oil transfer line between the
heater H1 and the column C1.

Filler materials used to create the welded joints were:

 Welding wire: CM5 –IG Boehler (WCrMo5Si), Ø2.40x1000 mm;
 Welding electrodes: Boehler Fox CM5–Kb (ECrMo5B42H5), Ø2.50x250 mm;

The welding process should have been realized with a PWHT.

An extensive experimental investigation has been carried out, including mechanical tests,
macro and microstructural examinations, chemical and electrochemical analyses on the specimens
extracted from the two coupons, according to the sampling plan presented in Figure 2.

a) Coupon containing a portion of the weld no.23 b) Coupon containing a portion of the weld no.24
between elbow 90 R=1.5D, Ø508×15.09mm and between elbow 90 R=1.5D, Ø508×15.09mm and
pipe Ø508×15.09mm pipe Ø508×15.09mm

Figure 2. Sampling plan for the specimens.

The chemical composition of the pipe, according to ASTM A335 P5, is shown in Table 1.
Int. J. Electrochem. Sci., Vol. 10, 2015 10836

Table 1. Chemical composition, manufacturer’s data [%]

Steel Grade C Mn P S Si Cr Mo
ASTM A335 Max. Max. Max. Max.
0.30 – 0.60 4.00 – 6.00 0.45 – 0.65
Grade P5 0.15 0.025 0.025 0.50

The main reference mechanical properties of the material, according to ASTM A335 P5, are
presented in Table 2.

Table 2. Mechanical properties (on the longitudinal direction)

Hardness
Tensile characteristics
characteristics
Steel Grade Yield strength Tensile strength Elongation at
Vickers
at +20°C at +20°C break at +20°C
[HV]
Rp0,2 [N/mm2] Rm [N/mm2] A5 [%]
ASTM A335
Min. 205 Min. 415 Min. 30 Max. 265
Grade P5

2.2 Experimental setup

In order to evaluate the current mechanical and structural characteristics of the samples taken
from the transfer line Ø508x15.09 mm, universal testing machine and optical microscopes were use
such as: universal testing machine EDZ 40 and ZD10/90, Charpy pendulum PSW 30, hardness
apparatus ZWICK 3212, stereo microscope MAK-MS, optical microscope MeF2.
Electrochemical tests were performed by using Autolab 302N EcoChemie after submerging the
iron electrode for 1h in 3% NaCl solution. OCP measurements were carried out at the Open-Circuit
Potential (OCP), the curves were recorded starting from the first moment of specimens sinking in a
free saline aerated test solution to observe the potential behaviour and qualitatively appreciate the
dissolution and/or creation of a protective film on the iron surface.
Potentiodynamic polarization testing allowed to calculate the corrosion parameters such as
corrosion potential (Ecorr), corrosion current (Jcorr), polarization resistance (Rp) and corrosion rate
(Rcorr) for specimens. The polarization curves were achieved by scanning the potential of iron from -
1200 to 200 mV at a scan rate of 1 mV/s. EW the equivalent weight in grams/equivalent of iron alloy
(EW = 27.93 grams/equivalent), d the density in g/cm3 (=7.86), and A the area of the exposed surface
of the electrode in cm2 (0.785 cm2). The polarization resistance, Rp (Ω), was achieved from the slope
of the plots of the applied potential against the measured current and the corrosion current Icorr (A) is
calculated using the Stern-Geary relationship with equation (1):

Icorr = βax βc /[ Rp 2.3(βa +βc)] (1)

Int. J. Electrochem. Sci., Vol. 10, 2015 10837

where: Tafel constants βa and βc represents the slope of the linear regions in V/decade of
current.

3. RESULTS AND DISCUSSION

Gravimetric methods were used to determine the chemical composition of specimens extracted
from the pipeline and welding zone. The specimen’s notation and obtained results are given in Table 3.
The chemical composition of specimens was determined similarly to the reference
specifications. By analyzing the chemical composition of the WM, it is noted that the WM is
compatible with materials of pipes and elbows.

Table 3. Chemical composition of the main components of the specimens

Specimen Mn Cr Mo Ni W Ti Cu Bi
1 Base material pipeline BM 0.38 4.80 0.45 - 0.61 - 0.08 0.33
2 Heat affected zone HAZe 0.32 5.10 0.48 0.10 0.38 - - 0.33
(elbow side)
3 Welded metal WM 0.47 5.13 0.52 0.10 - 0.37 - 0.10
4 Heat affected zone HAZp 0.48 5.00 0.46 0.01 0.62 - - 0.32
(pipe side)

The macroscopic analysis has revealed that the inner surface of the component is degraded by
selective corrosion phenomenon localized to the welded joint and adjacent areas, Figure 3. The
presence of active areas (areas with metallic gloss and colour change by corrosion products), and
passivated areas (dark) has been found. It appears that degradation was initiated in the weld (Figure 3a)
and has propagated to the BM of the two components. Also, in the BM, pitting degradation has been
identified (Figure 3b).

a) Weld no.24 – WM and HAZ areas (detail) b) Weld no.24 – HAZ and BM areas (detail)

Figure 3. Macroscopic images of WELD no. 24 (inner surface).

Int. J. Electrochem. Sci., Vol. 10, 2015 10838

The elbow BM (Figure 4a) has a ferritic structure with punctiform complex carbides and
non-metallic inclusions (oxides, silicates) with a grain size of 8-9. BM of the pipe (Figure 4b) has a
ferritic structure with complex carbides with a grain size of 6-7.

a) BM - elbow b) BM - pipe

Figure 4. Microscopic images of pipeline’s BM.

Evidenced structures are normal for these materials. The microstructures of HAZ and WM
zones from pipeline before and after PWHT are presented in Figures 5a -f. It was observed that the
structures before PWHT are not homogenous and presents rough seeds with a rather fibrous
morphology (Figures 5a, 5c and 5d).
Optical micrographs of HAZ elbow and side show pearlite-ferrite structures with complex
carbide and localized areas of bainite and martensite. The micrographs of weld sample show a
dendritic structure (pearlite, ferrite, complex carbide, bainite and martensite in localized areas) and
revealed in the ferrite part the presence of uniform polygonal ferrite seeds and the formation of some
globular carbide particles between the ferrite seeds. This could be connected with precipitation of
chromium carbide during the thermal treatment and be a source of inter-granular corrosion. Because
the Cr content in the sample was the same as in BM, the corrosion is rather a result of the lack of
PWHT (or improper realization on site at the elbow replacement).

a) HAZe– elbow side b) HAZe – elbow side (after PWHT)

Int. J. Electrochem. Sci., Vol. 10, 2015 10839

c) WM d) WM (after PWHT)

e) HAZp – pipe side f) HAZp – pipe side (after PWHT)

Figure 5. Influence of PWHT at the microstructural level. The indices “e” and “p” signify the elbow or
pipe

Very high (unaccepted) values of hardness in the WM and HAZe were measured, a maximum
of 405HV10 in WM and 397HV10 in HAZe has been determined. The material brittleness depends on
the martensite quantity, thus an optimum quantity is required because too much martensite makes the
material brittle and too little yields a soft material.
These measured values are over two times higher than the highest value of BM (199HV10) and
the increase of hardness is a result of the increase of martensite content over the optimum quantity, due
to the welding process. This demonstrates that in this circumstance the welding induced a sharp local
stress state. The local hardening estimator calculated with equation (2) demonstrates that, at the
boundary, the conditions are met for brittle fracture of the joint (ΔHV10> 50%).

HV10max  HV10min 405  199

ΔHV10 = 100  100 = 50.9 % (2)
HV10max 405

By varying heat treatment, improvement in the hardness of the specimens can be attained,
mainly depending upon the amount of martensite in the structure.
Int. J. Electrochem. Sci., Vol. 10, 2015 10840

In order to emphasize the influence of a correctly carried out PWHT, the sample was subjected
to heat treatment by heating up to + 725 °C with 60 minutes holding and controlled cooling at a rate of
100 °C/h up to 300 °C, followed by a free cooling in air (Figure 6). After the PWHT more
homogenous structures than those before heat treatment have been observed, with an island
distribution of martensite in a ferrite matrix (Figure 5f). The HAZe and HAZp of the elbow and pipe
side present a pearlite-ferritic structure, with sorbite areas, and the WM a fine dendritic structure
(pearlite, ferrite and sorbite).

Figure 6. Schematic illustration of the PWHT cycle.

By applying of PWHT, the maximum hardness of the WM was reduced up to 268HV10 and in
HAZp up to 266 HV10 and the maximum value determined for BM was 192HV10. In this case, the
local hardening estimator (relative to the BM) is (3):
268  192
ΔHV10 = 100  28.4% (3)
268
It is noted that the local tension induced by welding and the likelihood of brittle fracture has
been considerably reduced. The results of the local hardening estimator are correlated with the values
obtained by notch impact tests, as follows:
 Although for this material the notch impact characteristics are not guaranteed, at the
notch impact tests carried out it appears that the BM of the pipe and the elbow has a very good
toughness property, the fracture energy being 278-294 J for BM of pipes and 277-292 J for BM of
elbows.
 Unlike the BM, in case of WM and HAZe, toughness characteristics have been
drastically reduced, to 80 – 228 J for HAZ and only 19 – 39 J for WM. This demonstrates that the WM
is fragile.
The changes of morphologies and mechanical properties induced by welding and heat
treatment are reflected in the electrochemical behaviour of the specimens. Corrosion measurements
can be used in evaluation of the quality of welding and to appreciate if the weld material and welding
process guarantee a service without failures. The variation of OCP potential in time for the specimens
immersed in 3% NaCl solution are presented in Figure 7.
Int. J. Electrochem. Sci., Vol. 10, 2015 10841

The potential of pipelines specimens in chloride solution are shifted toward the negative
direction in the first minutes of iron immersion. The potential shift is due to the dissolution of iron as
aggressive chloride ions attack the surface. In time the corrosion products formed a layer on the
surface and partially protected the iron from further attack.

-0.30

-0.35

BM
-0.40
WM
E, V (Ag/AgCl)

HAZC
-0.45 HAZT

-0.50

-0.55

-0.60

0 500 1000 1500 2000 2500 3000 3500 4000

Time, min.

Figure 7. OCP curves for specimens recorded in 3.0% NaCl solution.

As a consequence, in time the slight shift in the potential towards the less negative values
appears. It was observed that the WM specimen showed a rapid negative shift in the first few minutes
and the highest shifts of the potential toward the negative value comparatively with the BM, HAZ e and
HAZp specimens. This means that WM is more prone to chloride attach, due to modification of
structure of the surface. Figure 8 shows Tafel polarization curves for pipeline specimens BM, WM,
HAZe, and HAZp after 90 min immersion in 3.0 wt % NaCl solutions.

0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
2
Log J, A/cm

-3.0
BM
-3.5
WM
-4.0
HAZC
-4.5
HAZT
-5.0
-5.5
-6.0
-6.5
-7.0
-7.5
-8.0

-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
E, V (Ag/AgCl)

Figure 8. Polarization curves for pipeline samples BM, WM, HAZe and HAZp after 90 min immersion
in 3.0 wt % NaCl solutions.
Int. J. Electrochem. Sci., Vol. 10, 2015 10842

Parameters obtained from polarization curves presented in Figure 8 are shown in Table 4.

Table 4. Electrochemical parameters for pipeline samples BM, WM, HAZe, and HAZp after 90 min
immersion in 3.0 wt % NaCl solutions at 25 °C.

Jcorr, Rcorr.
Sample c, V a, V Rp, Ohm, Ecorr, Ohm
A/cm2 mm/year
BM 1.654x10-6 0.076 0.126 3.208x103 -0.860 0.019
WM 2.638x10-5 0.163 0.074 2.534x102 -0.523 0.307
HAZp 9.571x10-7 0.094 0.101 5.462x103 -0.751 0.011
HAZe 1.676x10-5 0.127 0.082 3.421x102 -0.810 0.195

The values obtained for Ecorr and Jcorr indicate their dependence on the structural
modification appeared due to the welding process. Post heat treatment improves mechanical properties
but could decrease the corrosion resistance. It is reported for low carbon steel that iced-water reduce
procedure generates a fine grained recrystallized structure and enhances the mechanical properties and
on the other hand, corrosion resistance decreases [19]. All specimens present a pearlite-ferrite structure
and the dissimilarity is a result of the difference in the microstructure (different spacing of the
cementite lamellae-ferrite, content and size of ferritic phase). The corrosion potentials determined for
BM, HAZp, WM are more appropriate to -0.860V (to base material), except the HAZe specimen which
present a nobler corrosion potential of –0.523 V. The nobler corrosion potential value obtained for
WM suggests that this zone is more prone to cover with a protective layer, comparatively with BM,
HAZp and HAZe specimens due to the formation of martensitic phase. It is reported that due to the
carbon content martensitic phase frequently shift the corrosion potential to a nobler value than the
ferritic matrix [14, 20, 22]. Higher hardness obtained for WM and HAZe specimens is also a
consequence of the formation of a martensite region. The presence of martensite in steel reduces the
corrosion resistance as was observed for WM and HAZe specimens. The Jcorr determined from Tafel
slope show higher corrosion current density for WM specimens and HAZe and lower corrosion current
density HAZp and BM. The welding induced in the specimens WM and HAZe appropriate conditions
to brittle fracture of the joint and is the reason for the higher Jcorr value obtained for these specimens.
The WM shows a dendritic structure (pearlite, ferrite, complex carbide, bainite and martensite in
localized areas) capable in an easier manner to incite an increase of the corrosion current density. The
increase of corrosion current density is 16 times higher comparatively with BM. The pearlite-ferrite
structure with complex carbide and localized areas of bainite and martensite observed for HAZ e
structure show a corrosion current 10 times higher comparatively with BM. Also, in case of WM and
HAZe the obtained higher corrosion currents and reduced toughness characteristics are the result of
residual stress appeared due to martensite transformation (less homogeneous microstructure); and due
to the fine needles or lamellas interconnection path between the martensite and ferrite phases (active
corrosion area). This significantly affects the ratio between the nobler and active phases responsible for
corrosion. Similar observation was reported in the literature [14]. The cathodic βc constant for BM,
Int. J. Electrochem. Sci., Vol. 10, 2015 10843

HAZp are 0.076 and 0.094V/decade. HAZe and WM specimens present higher values and this is a
result in variable degradation due to breakdown of the passivating film, behaviour characteristic for
alloy materials indicating differential Redox reactions. For HAZe and WM the cathodic Tafel slopes
(βc) are found to be greater than those anodic Tafel slopes (βa) and indicate an overall kinetics under
cathodic control due to the increased currents on values of anodic part.
After PWHT, the HAZp specimen shows reduced local stress state and this is reflected in the
value for Ecorr and Jcorr due to increasing in homogeneity of structure. In result, a decrease of Jcorr of
HAZp with about 17 times lower comparatively with HAZe was obtained. Low value for corrosion rate
Rcorr are found for HAZp and BM specimens.
The microstructural differences are revealing in the EIS spectra. EIS tests were performed in a
3% NaCl solution at 25 °C. Impedance spectra were illustrated in both complex impedance diagrams
(Nyquist plot), and Bode amplitude and phase angle plots. In the Nyquist graph (Figure 9a), the
imaginary component of the impedance is represented as a function of the real component, whereas the
Bode plots (Figure 9b) illustrates the logarithm of the impedance modulus |Z| and phase angle as a
function of the frequency f logarithm.

BM
600 3.0 50
WM BM
HAZC WM

-phase, degree (empty symbols)

500 HAZT HAZC
40
2.5 HAZT
logIZI, Ohm (fill symbols)

400
30
2.0
Zim, ohm

300
20
1.5
200
BM 10
WM
1.0
100 HAZC
HAZT 0

0 0.5
0 200 400 600 800 1000 1200 -2 -1 0 1 2 3 4 5
Zre, ohm log f, Hz

a) b)

Figure 9. a) Complex Nyquist plane plots and b) Bode plots for the studied specimens.

Electrochemical parameters were extracted using an electric equivalent circuit (EEC) shown in
Figure 10.

Figure 10. Schematic representation of the equivalent electric circuit.

Int. J. Electrochem. Sci., Vol. 10, 2015 10844

The EEC was chosen to fit the data, it included an electrolyte resistance – Rs which is
approximately constant in all determinations, the polarization resistance - Rp of the film formed on
each specimen surface which represents the coating film resistance as a result of formation of a oxide
layer and/or corrosion products and corresponds to the corrosion resistance and Q represents the
capacitance of the film. Impedance parameters obtained from fitting the impedance data were shown in
Table 5.

Table 5. Values of the electric circuit elements for the studied specimens.

Rs Q-T Q-P, Rp
Specimen Chi-Sqr Sum-Sqr .cm2 F/cm2/sф−1 (ф) .cm2
BM 2.76E-03 9.755 9.755 1.97E-03 0.631 1696
WM 1.35E-02 9.706 9.706 4.11E-03 0.630 324.3
HAZe 4.37E-03 4.603 4.603 1.17E-02 0.600 1276
HAZp 7.94E-03 6.034 6.034 2.88E-03 0.490 6882

The Q values are correlated with the local dielectric constant and/or thickness of the electrical
double layer variation. The capacitance value could decrease due to an increase of protective layer
thickness or to decrease in dielectric constant of protective layer. Higher Q values are obtained for
HAZe and WM specimens and indicate a decrease in thickness of the protective layer.
An increase for the parameter Q-T and a decrease for Rp value the case of HAZe and WM
pointed out a passive layer with more defects or less thickness. The values of Rp and Q-T for the
HAZp suggest a thicker surface or with enhanced density with fewer defects.
The results reveal that the Rp value also depends on the microstructural patterns. The Rp is
affected by the thickness and composition of the passive layer formed on surfaces and affect their
corrosion resistance. Electrochemical parameters indicate that, the WM presents lower Rp value. The
Rp value of WM pointed to the formation of a passive layer with more defects or less thickness than
the BM, due to the microstructural changes produced by the thermal flux introduced into the welding
area (unproper PWHT). The microstructural change could be associated with the apearance of
martensitic phase, and the martensitic transformation is possible to be less uniform in the whole part
(Figure 5a, c and d). Similar observation was reported for austenitic stainless steel deformation -
induced α’ martensitic transformation arises by plastic deformation, which does not occur in the whole
part [21, 22]. By changing the PWHT an increase by 5.4 times in the Rp was achieved as can be seen
for HAZp comparatively with BM and HAZe specimens, pointing out an increase tendency to cover
with a protective oxide film, due to the formation of martensitic phase in an optimum quantity and
grain size. These results confirm that the proper PWHT leads to a high resistance to the weld-decay.
The results of potentiodynamic polarization further support the results of electrochemical impedance
spectroscopy and also indicate that the proper PWHT determines a proper corrosion resistance of the
welded material.
Int. J. Electrochem. Sci., Vol. 10, 2015 10845

4. CONCLUSIONS

The paper present the importance of convergence of the mechanical testing results with those
of microstructural analysis, chemical and electrochemical evaluation, to achieve relevant information
on corrosion behaviour of welded joints, which could lead to the cause of pipeline failures, related to
the actual material degradation condition and technological process parameters: temperature, loading
conditions and working environment.
It is demonstrates that hardened steel (consisting of ferrite and martensite and carbides) corrode
much faster than a material in a state of balance (normalized). The results of the performed
experimental program lead to the conclusion that preferential corrosion of pipeline at the WM and the
HAZ was possible due to the metallurgical effect and the stress concentration in the adjacent area of
the weld, due to the presence in the fluid of corrosive compounds, under normal operating conditions
of the system at elevated temperatures. The metallurgical effect refers to the significant alteration of
the corrosion rate of the intrinsic material, due to the microstructural changes produced by the thermal
flux introduced into the welding area (inappropriate linear energy using, high cooling rate, the absence
of a PWHT, or incorrect PWHT performed).

References

1. W. T. Becker and J. R. Shipley, ASM Handbook Volume 11: Failure Analysis and Prevention,
ASM International, Ohio (2002)
2. P. Marcus, ASM Handbook Volume 13A: Corrosion: Fundamentals, Testing and Protection, ASM
International, Ohio (2003)
3. M. Alizadeh and S. Bordbar, Corros. Sci., 70 (2013) 170
4. C. W. Du, X. G. Li, P. Liang, Z. Y. Liu, G. F. Jia and Y. F. Cheng, J. Mater. Eng. Perform., 18
(2009) 216
5. G. Zhang and Y. Cheng, Corros. Sci., 51 (2009) 1714
6. S. Bond, ASM Handbook Volume 13A: Corrosion: Fundamentals, Testing and Protection, ASM
International, Ohio (2003)
7. E. Sadeghi Meresht, T. Shahrabi Farahani and J. Neshati, Corros. Sci., 54 (2012) 36
8. H. Kokawa, M. Shimada, M. Michiuchi, Z. J. Wang and Y. S. Sato, Acta Mater., 55 (2007) 5401
9. S. J. Oh, D. C. Cook and H. E. Townsend, Hyperfine Interact., 112 (1998) 59
10. C. Du, X. Li, P. Liang, Z. Liu, G. Jia and Y. Cheng, J. Mater. Eng. Perform., 18 (2009) 216
11. E.-S. M. Sherif, R. M. Erasmus and J. D. Comins, Electrochim. Acta, 55 (2010) 3657
12. E.-S. M. Sherif, Molecules, 19 (2014) 9962
13. Z. Zhang, S. Chen, Y. Li, S. Li and L. Wang, Corros. Sci., 51 (2009) 291
14. P. P. Sarkar, P. Kumar, M. K. Manna and P. C. Chakraborti, Mater. Lett., 59 (2005) 2488
15. W. R. Osório, L. R. Garcia, P. R. Goulart and A. Garcia, Mater. Chem. Phys., 106 ( 2007) 343
16. R. W. Cahn, Encyclopedia of Materials Science and Engineering Volume 2, Pergamon Press,
Elmsford New York (1986)
17. P. R. Roberge, Handbook of Corrosion Engineering, McGraw-Hill, Professional Publishing, New
York (2000)
18. API Recommended Practice 571, Damage Mechanisms Affecting Fixed Equipment in the Refining
Industry, API Publishing Services, Washington DC (2011)
19. A. I. Almarshad and S. A. Hossain, Corros. Eng., Sci. Technol., 5 (2006) 7
20. W. R. Osorio, L. C. Peixoto, L. R. Garcia and A. Garcia, Mater. Corros., 60 (2009) 804
Int. J. Electrochem. Sci., Vol. 10, 2015 10846

21. N. Solomon and I. Solomon, U.P.B. Sci. Bull. Series B, 72 (2010) 1454
22. Y. T. Zhao, S. W. Yang, C. J. Shang, X. M. Wang, W. Liu and X. L. He, Mater. Sci. Eng., 454–
455 (2007) 695

© 2015 The Authors. Published by ESG (www.electrochemsci.org). This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/4.0/).