Materials Sciences and Applicatio n, 2011, 2, 917-921
doi:10.4236/msa.2011.27122 Published Online July 2011 (
Copyright © 2011 SciRes. MSA
Influence of S Contents on the Hydrogen Blistering
and Hydrogen Induced Cracking of A350LF2 Steel
Shuqi Zheng*, Changfeng Chen, Liqiang Chen
Department of Materials Science and Engineering, China University of Petroleum-Beijing, Beijing, China.
Email: *
Received March 25th, 2011; revised April 20th, 2011; accepted May 3rd, 2011.
In the present work, the effects of chemical compositions on the hydrogen blistering and hydrogen induced cracking of
A350LF2 steel with different S contents were studied. Four types of A350LF2 steels were evaluated by immersing sam-
ples in H2S-saturated NACE solution. The hydrogen blistering, crack length rate(CLR), crack thickness rate(CTR) and
crack sensitivity rate(CSR) were evaluated. The results show that there are many hydrogen blisters on the sample sur-
face with S content of 0.021%, a few on the sample surface with S content of 0.019% and 0.012% and no one on the
surface with S content of 0.002%. There were 12, 2 and 1 strips of cracks of longer than 0.3 mm on the evaluated cross
sections with S content of 0.021%, 0.019% and 0.012%, respectively. There was no any crack in the sample with
0.002% S. The corrosion rate was also evaluated. The S content has no obvious influence on the corrosion rate.
Keywords: A350LF2, S Content, Hydrogen Blistering, Hydrogen Induced Cracking
1. Introduction
At present, more than four hundred sour oil and gas
fields containing hydrogen sulfide (H2S) have been
discovered in the world, in which the hydrogen sulfide
content vary greatly from micro-content H2S to above
92% [1]. For exploitation and transportation of oil and
gas which contain hydrogen sulfide (H2S) safely, the
choice of oilfield equipments and pipeline steels is
strict due to mechanical and corrosion problems in-
duced by hydrogen permeation.
When oilfield equipments are exposed to wet hy-
drogen sulfide environment, hydrogen atoms are pro-
duced by the surface corrosion of the steel. In the
presence of hydrogen sulfide, recombination reaction
of hydrogen atoms to the molecular hydrogen is re-
tarded, consequently allowing, hydrogen atoms to dif-
fuse into the ste el. These atoms can move rap idly e ither
by diffusion or by transportation through mobile line
defects [2], diffused hydrogen atoms are trapped in
sensitive metallurgical defects such as the interface
between non-metallic inclusion and steel matrix.
Cracking can occur if the critical amount of hydrogen
necessary for crack initiation is accumulated [3,4].
According to the different formation of the crack, it is
divided into three main types, such as hydrogen in-
duced cracking (HIC), sulfide stress cracking (SSC)
and Hydrogen blistering (HB). HIC and HB can result
in premature failure of equipment. Some researchers
believe HIC and HB are caused by inclusion distribu-
tion, especially by the elongated MnS inclusion distri-
bution [5,6]. But the relationship between chemical
compositions, especially the S content and cracking
sensitivity has not been studied systematically in
A350LF2 steels and the influence of S content on the
corrosion rate is not very clear. In this paper, the ef-
fects of chemical compositions of four types of
A350LF2 steels with different sulfur content on crack-
ing and corrosion rate are investigated.
2. Experimental Procedures
2.1. Specimens and HIC Test
The materials used in the test were A350LF2 flange
steels with different S contents of 0.021%, 0.019%,
0.012% and 0.0002%, respectively. The testing of the
steels’ resistance to HIC were performed in NACE solu-
tion A (5% NaCl plus 0.5% acetic acid solution saturated
with H2S) at 25˚C for 96 h according to NACE TM0284
-2003 standard [7 ].
Three specimens (100 mm× 20 mm× 30 mm) were
tested for each flange. Before the HIC test, the surfaces
were ground with SiC abrasive paper down to 800 mesh.
After exposure, all immersed specimens were cut and
Influence of S Contents on the Hydrogen Blistering and Hydrogen Induced Cracking of A350LF2 Steel
polished and then the cross-sections were examined with
100X optical microscope (OM) after etched with solution
of a mixture of 4% nitric acid and ethanol. Three differ-
ent cracking parameters were measured:
Crack Length Ratio (CLR) = a/W × 100% (1)
Crack Thickness Ratio (CTR) = b/T × 100% (2)
Crack Sensitivity Ratio (CSR) = (a × b)/(W × T) × 100%
where a is the crack length, b is the crack thickness, W is
the section width and T is the test specimen thickness.
2.2. H2S Corrosion Rate Test
Corrosion tests were carried out in the H2S corrosion
experimental device following the NACE Standard
TM-0284-2003. The size of corrosion rate sample is 50
mm × 10 mm × 3 mm. Before the test, the surface of the
test specimens were polished with grit silicon carbide
papers progressively up to 800 grade, then degreased
with acetone and rinsed with absolute alcohol, weighted
using a precision of 0.001 mg, and finally stored in des-
iccators for use. Before experiment, the sealed test vessel
was deoxygenated with nitrogen for at least 6 hours.
Purging began immediately after the test vessel was
filled and was done at a rate of at least 100 ml per minute,
per liter of test solution. After purging, H2S gas was be
bubbled through the test solution. The rate of bubbling
was at least 200 ml per minute, per liter of test solution
for the first 60 minutes. Thereafter, a positive pressure of
H2S gas was maintained which controlled by the
flowmeter. After 168 hours’ corrosion, the samples were
removed from the solution and rinsed with deionized wa-
ter. The samples were descaled (the solution: 1 L HCl (ρ
= 1.19 g/L), 20 g Sb2O3 and 50 g SnCl2), rinsed with wa-
ter and absolute alcohol, dried in nature state and weight-
ed again with a precision of 0.001 mg. The corrosion rate
was represented by corrosion depth (mm)/corrosion time
(per year), i.e. mm/a. The corrosion morphologies were
observed by means of scanning electronic microscopy .
3. Results and Discussion
The surface morphologies of the flange samples after 96
hours’ exposure in NACE A solution are shown in Fig-
ure 1. From the figure, we found that the sensitivity to
hydrogen blistering for four types of samples is far dif-
ferent. It is known from the previous research that the S
concentration is the most important factor of improving
HIC resistance in low alloy steel. Many HBs were ob-
served uniformly distributed on the sample surface with
S concentration of 0.021% (Figure 1(a)). More than 316
hydrogen blisters was counted on the observed surfaces
of a group of samples which were composed of 3 speci-
mens by the naked eye observation. The area of hydro-
(a) (b)
(c) (d)
Figure 1. The hydrogen blistering on the sample surfaces
with different S concentration after HIC test (a) 0.021% S;
(b) 0.019% S; (c) 0.012% S; (d) 0.002% S.
gen blisters was about 8.9% of surface area. There were
61 hydrogen blisters with the diameter of bigger than 1
mm. By comparison, the amount of hydrogen blisters on
the sample surface with S concentration of 0.019% de-
creased obviously with 37 HBs on the observed surface
and about 1.2% sample surface area was occupied by
HBs. There were 26 HBs with the diameter of more than
1 mm. But HBs were not observed uniformly distributed
on the sample surface but most HBs were located on one
sample surface. There were only 3 HBs on the observed
surface with S concentration of 0.012% and the HBs area
rate was less than 0.1% of the observed sample surface
area. However, no HB was found on the surface with S
concentr ation of 0.002%. So we can draw the conclusion
that the hydrogen blisters depend on the S content of
A350LF2 steel. When the S concentration is more than
0.02%, the amount and area of HB increased obviously
while when the S content is less than 0.02%, the amount
and area of HB decreased obviously, and when the S
content tends to 0.002%, no HB was observed on the
sample surface.
To study the HB, One of hydrogen blisters was torn
off. The scanning electronics microscopy (SEM) of the
HB bottom was shown in Figure 2. It was shown that
there were many inclusions at the bottom of HB. It indi-
cates that the nucleation site of a HB is just in the posi-
tion of inclusion. In the HIC test, Since H2S,a hydrogen
promotor, acts as a catalytic poison, just a small part of
hydrogen recombines or associates on the surface and
then bubbles off, yet most hydrogen atoms penetrate into
specimens, but that isn’t enough to induce hydrogen
pressure. Only in the hydrogen-cavity or micro-cracks
can the hydrogen atoms combine into hydrogen mole-
cules whose aggregation can form hydrogen pressure.
The hydrogen formed here ultimately can not bubble off.
Copyright © 2011 SciRes. MSA
Influence of S Contents on the Hydrogen Blistering and Hydrogen Induced Cracking of A350LF2 Steel919
Figure 2. SEM morphology of the bottom of a hydrogen
Owing to this, the hydrogen pressure increases and when
the partial pressure exceeds the strength of the material,
hydrogen blistering occurs.
Hydrogen absorbed from the solution diffuses in the
metal and then recombines as hydrogen molecules at trap
sites in the steel matrix (favorable trap sites are typically
found in rolled products along elongated inclusions or
segregated bands of microstructure). High hydrogen
pressure increases along with the initiation of cracks. As
more hydrogen enters the voids, the pressure rises, de-
forming the surrounding steel so that blisters may be-
come visible on the surface. The steel around the crack
becomes highly strained and this can cause linking of
adjacent cracks to form stepwise cracking.
After testing, each test specimen was sectioned ac-
cording to NACE TM 0284-2003 and the cross section
surfaces were examined. Each section was polished met-
allographically and etched so that cracks can be distin-
guished from small inclusions, laminations, scratches, or
other discontinuities. In measuring crack length and
thickness, cracks separated within the distance of less
than 0.5 mm were considered a single crack. The crack
morphologies of different types of specimens were
shown in Figure 3. Figures 3(a) and (b) are the typical
cracks observed on each section surface of the samples
with S concentration of 0.021%. We found that some
cracks were very long while some were very short. The
stepwise propagation of HIC, also known as stepwise
cracking, is shown in Figure 3(a) which were connected
by some cracks on different layers. There were also
many discontinuous cracks and the cracks assumed in
different directions. There were about 12 cracks which
were longer than 0.3 mm on all observed section faces.
Figure 3(c) is the typical crack character of the sample
with S content of 0.019%. The crack morphologies were
the same as those of the sample with S content of 0.021%.
There were 2 cracks wh ich were longer than 0.3 mm and
9 cracks longer than 0.15 mm.
In comparison there were only some very short cracks
(a) (b)
(c) (d)
Figure 3. The typical hydrogen induced cracks in the cross
section of A350LF2 flange with different S contents (a)(b)
0.021%; (c) 0.019%; (d) 0.012%.
observed in the sample with S content of 0.012% (Figure
3(d)), and no crack was found in the sample with S con-
tent of 0.002%. So, th e amount an d leng th of cracks ha s a
dependence on S content of the samples.
According to Equations (1), (2) and (3), the value of
CLR, CTR and CSR for each sample are calculated. The
CLR value less than 15% was adopted as an acceptance
criteri on of HIC test according to ISO 15156.2 [8]. There
are some other acceptance criterions of HIC quality which
lie on the served environments and the materials users [9].
It was found that only CLR of flange 1 exceeded 15% and
the other 3 flanges can also be accepted. But with the ex-
ploration and exploitation of oil and gas fields containing
high partial pressure H2S, disasters happen sometimes. In
order to avoid the occurrence of disaster, the stricter ac-
ceptance criterions were required. So the endeavor to im-
prove the HIC resistance of flange is one of urgent affairs
for materials researchers. The values of CLR, CTR and
CSR of different flanges were shown in Figure 4. Figure
4(a) shows the CLR value of four types of flanges. It was
shown that there is obvious difference between S content
more than 0.02% and less than 0.02%. And we also
found that for one flange samples, the values of each
section were also different. This is the reason why at
least 3 samples requested for HIC test. The plot of aver-
age values of CTR and CSR of different samples were
shown in Figures 4(b) and (c). The same tendency was
found for CTR and CSR. So the flange with more than
0.02% of S content should better not be used in wet H2S
environment, the flange with S content from 0.01% to
0.02% should better be used in less severe environment,
and in the severe environments S content of the flange
should be controlled less than 0 .01%.
Figure 5 shows the average corrosion rates (CR) of
different flange samples after 168 hours’ exposure in
saturated NACE A solution. The result from Figure 5
indicates that the average corrosion rates are about
0.8mm/a for each sample with different S contents,
Copyright © 2011 SciRes. MSA
Influence of S Contents on the Hydrogen Blistering and Hydrogen Induced Cracking of A350LF2 Steel
0.000 0.005 0.0100.015 0.020
Corrosion R ate(mm/a)
S content (wt%)
0.000 0.005 0.010 0.015 0.020
CLR (%)
S content (wt%)
Sample 1
Sample 2
Sample 3
no crackshort cracklong crack
0.000 0.005 0.010 0.015 0.020
CTR (%)
S content (wt%)
long crack
short crack
no crack
0.000 0.005 0.010 0.015 0.020
CSR (%)
S content (wt%)
no crackshort crack
Figure 4. CLR, CTR and CSR of the A350LF2 flanges with
different S content.
which means that the corrosion rate has not an obvious
relation with the S content of flanges.
Figure 6 shows the morphologies of the scale formed
on different flange samples. It is found that they are all
covered with a layer of corrosion film on the sample sur-
faces. The EDS results confirm that the key component is
iron sulfide. By comparison, it is discovered that there
are no large differences between the corrosion products
on the surfaces of the samples with different S content,
which proves that S content has slight effect on the cor-
rosion film and corrosion rate.
4. Conclusions
The effect of chemical compositions of four types of
Figure 5. The corrosion rate of the specimens with different
S content.
Figure 6. The morphologies of corrosion scales formed on
the flange samples with different S content. (a) 0.021%; (b)
0.019%; (c) 0.002%.
Copyright © 2011 SciRes. MSA
Influence of S Contents on the Hydrogen Blistering and Hydrogen Induced Cracking of A350LF2 Steel
Copyright © 2011 SciRes. MSA
A350LF2 steels with different S con tent on cracking and
corrosion rate has been carried out. The results are sum-
marized as follows:
1) The chemical compositions of A350LF2 flange
steel, especially S content have a strong impact on the
hydrogen blistering and hydrogen induced cracking;
2) As for A350LF2 flange steel, with the rise of S
content, the sensitivity to the hydrogen blistering and
hydrogen induced cracking gradually increases. When
the S content is more than 0.02%, the hydrogen blistering
and hydrogen induced cracking are far beyond the stan-
3) The corrosion rate of A350LF2 steel is less affected
by S content.
5. Acknowledgements
This work was financially supported by the following
funds:) the Scientific Research Fund for Returned Over-
seas Chinese Scholar from Education Ministry of China
(No. 2008-890);) the Natural Science Foundation of
China (No. 50871122)
[1] G. Y. Zhu, J. X. Dai and S. C. Zhang, “Generation
Mechanism and Distribution Characters of Hydrogen
Sulfide Bearing Gas in China,” Natural Gas Geoscience,
Vol. 15, No. 2, 2004, pp. 166-171
[2] P. J. Ferreira, I. M. Robertson and H. K. Birnbaum,
“Hydrogen Effects on the Character of Dislocations in
High-Purity Aluminum,” Acta Materialia, Vol. 47, No.
10, 1999, pp. 2991-2998.
[3] G. T. Park, S. U. Koh, H. G. Jung and K. Y. Kim, “Effect
of Microstructure on the Hydrogen Rapping Efficiency
and Hydrogen Induced Cracking of Pipeline Steel,” Cor-
rosion Science, Vol. 50, No. 7, 2008, pp. 1865-1871.
[4] W. K. Kim, S. U. Koh, B. Y. Yang and K. Y. Kim, “Ef-
fect of Environmental and Metallurgical Factors on Hy-
drogen Induced Cracking of HSLA Steels,” Corrosion
Science, Vol. 50, No. 12, 2008, pp. 3336-3342.
[5] T. Hara, H. Asahi and H. Ogawa, “Conditions of Hydro-
gen-Induced Corrosion Occurrence of X65 Grade Line
Pipe Steels in Sour Environments,” Corrosion, Vol. 60,
No. 12, 2004, pp. 1113-1121. doi:10.5006/1.3299225
[6] M. A. Lucio-Garcia, J. G. Gonzalez-Rodriguez and M.
Casales, “Effec t of Heat Treatment on H2S Corrosion of a
Micro-Alloyed C-Mn Steel,” Corrosion Science, Vol. 51,
No. 10, 2009, pp. 2380-2386.
[7] NACE standard TM0284-03, “Evaluation of Pipeline and
Pressure Vessel Steels for Resistance to Hydrogen-In-
duced Cracking,” NACE International, Huston.
[8] International standard ISO 15156-2:2003, Petroleum and
Natural Ga s In dustries. Materials for Use in H2S-Containing
Environments in Oil and Gas Production -Part 2: Crack-
ing-Resistant Carbon and Low Alloy Steels, and the Use
of Cast Irons.
[9] N. Loukachenko, P. Bourges, K. E. Orie, C. Chauvy and
L. Coudreuse, “Recent Experience on Sour Service Re-
sistant Steels Behavior,” NACE Corrosion 2009, Paper
No. 09352.