CORROSION SCIENCE SECTION

Effect of Testing Temperature on Sulfide

Stress Cracking of Low Alloy Steel

Kenji Kobayashi,‡,* Tomohiko Omura,* and Masakatsu Ueda**

ABSTRACT KEY WORDS: hydrogen absorption, hydrogen embrittlement,

hydrogen permeation, sulfide stress cracking
In this study, effects of environmental temperature on sus-
ceptibility to sulfide stress cracking (SSC)—a type of hydrogen
embrittlement (HE) occurring in sour environments—of low INTRODUCTION
alloy steels were investigated from the perspective of hydrogen
entry, absorption, and accumulation. SSC susceptibility was In oil and gas fields, which contain hydrogen sulfide
evaluated using a double cantilever beam (DCB) test and a four- (H2S) with condensed water or formation water (the so-
point bend (4PB) test in sour environments at several testing called sour environments), a significant amount of
temperatures. 4PB test specimens included notched and un- hydrogen originates from a cathodic reaction, which
notched specimens to investigate influences of stress con- accompanies a corrosion reaction on steel and enters
centration and local stress. In the case of evaluation methods the steel through a poisonous catalytic action of H2S.
using specimens with high-stress concentration area, a de- It is well known that the hydrogen causes hydrogen
crease in testing temperature from room temperature to 4°C
embrittlement (HE) of steel, for example, sulfide stress
significantly increased SSC susceptibility. Hydrogen entry
cracking (SSC) and hydrogen induced cracking (HIC)
and absorption behaviors were also evaluated at several testing
temperatures using a hydrogen permeation test. The hydro- in sour environments. HE is the biggest barrier in
gen concentration at the plastic deformed area increased applying higher strength steels to oil and gas fields
remarkably with decreasing testing temperature. It is con- containing H2S because an increase in the strength level
sidered that the influence of testing temperature is due to of steels increases HE susceptibility. In particular,
hydrogen concentration at the stress concentration area with SSC is a major issue for high-strength low-alloy steel Oil
plastic deformation. In a low temperature condition, the degree Country Tubular Goods (OCTG), which is used for oil
of hydrogen accumulation at the crack tip areas of a DCB and gas production.
specimen or crack initiation site of a 4PB specimen could be SSC is caused by both hydrogen and tensile
higher than that in a higher temperature condition. When
stress. Several standard test methods to evaluate SSC
steels are applied to low temperature conditions with H2S,
susceptibility of steels have been published by the
a prior material evaluation reproducing both environmental
temperature and actual stress condition is needed.
National Association of Corrosion Engineers (NACE)
International and European Federation of Corrosion
(EFC). NACE TM0177 includes four test methods
for SSC or SCC evaluations: tensile test (method A),
Submitted for publication: July 28, 2017. Revised and accepted:
January 7, 2018. Preprint available online: January 9, 2018,
three-point bend test (method B), C-ring test (method C),
[Link] and double cantilever beam (DCB, method D) test.1
‡
Corresponding author. E-mail: [Link]@[Link]. A testing temperature is one of the most important
* Nippon Steel & Sumitomo Metal Corporation, R & D Laboratories,
1-8 Fuso-cho, Amagasaki, Japan. factors with which to appropriately evaluate the SSC
** Formerly at Nippon Steel & Sumitomo Metal Corporation. susceptibility of steels. Further consideration of the
ISSN 0010-9312 (print), 1938-159X (online)
CORROSION—Vol. 74, No. 6 18/000099/$5.00+$0.50/0 © 2018, NACE International 603
CORROSION SCIENCE SECTION

testing temperature is required by accounting for the investigate the influences of stress concentration and
diversification of the application condition, as mentioned local plastic deformation on the SSC test results in
below. At present, the testing temperature is fixed at detail. Hydrogen entry and absorption behaviors were
24°C (75°F) for SSC evaluation regardless of the evalu- also evaluated using hydrogen permeation tests. Then,
ation method in NACE TM0177. In EFC 16, it is fixed the mechanism of temperature dependence of SSC
at 23°C, while in ANSI/NACE MR0175/ISO 15156-2, a susceptibility is discussed from the perspectives of
material that passes at around 24°C is applicable at all hydrogen entry, absorption, and accumulation to the
temperatures.2–3 Since previous studies, e.g., by Town- crack initiation and propagation area.
send,4 have shown that SSC susceptibility at room
temperature is the most severe, room temperature
(around 24°C) has been recommended as the testing
EXPERIMENTAL PROCEDURES
temperature in these test methods. That study reported
the effect of testing temperature on SSC within a range
Material
A high-strength low-alloy steel OCTG with a
of 0 to 80°C using a bent wire specimen. Bent specimens
specified minimum yield strength of 758 MPa (110 ksi)
tested at around 25°C showed the shortest time-to-
for sour service produced by quenching and tem-
failure compared to those at lower or higher tempera-
pering heat treatments was used in this study. The
tures. However, the most important thing is not the
chemical compositions and mechanical properties
time-to-failure but whether SSC can occur at the envi-
are listed in Tables 1 and 2, respectively. The material
ronment or not. At low temperature, the time-to-
exhibited tempered martensitic microstructure.
failure may be longer than that at high temperature
because the hydrogen diffusion coefficient decreases
with decreasing the testing temperature. NACE TM0177 Test Solution
indicates that an SSC test below 24°C increases SSC Two test solutions shown in Table 3 were used for
severity. However, a detailed examination of the effect of the sour corrosion tests in this study. Solution A was an
testing temperature on SSC susceptibility, especially aqueous solution with 5% NaCl and 0.5% CH3COOH
at temperatures lower than room temperature, has not (HAc) saturated with 0.1 MPa of H2S. It corresponds to
been accomplished. NACE TM0177 solution A. Solution B′ was defined as
In recent years, the application of low-alloy steel an aqueous solution with 5% NaCl, 0.4% CH3COONa
OCTG in an environment with temperature lower than (NaAc), and CH3COOH saturated with 0.1 MPa of
room temperature has increased. With a drain on H2S. The initial pH was adjusted to 3.5 through
onshore oil and gas wells, offshore fields have been CH3COOH addition. Solution B′ was similar to NACE
actively developed. In these fields, some of the TM0177-2016 solution B. The buffering capacity of
equipment are exposed to environments with relatively solution B′ was higher than that of solution A because
low temperatures representative of seawater tem- of the increments in HAc and NaAc. Before saturation
perature at the seabed (around 4°C).5 SSC evaluation with H2S, the test solution and a test vessel were
in a low-temperature environment has become an purged with N2 to remove the oxygen contamination.
important feature for the steels used in offshore oil After that, H2S was injected into the test vessel. The
and gas fields. There have been some recent reports
on the effect of testing temperature on SSC using TABLE 1
DCB tests.6-11 Miglin, et al., evaluated the effect of Chemical Compositions of Material/wt%
testing temperature on DCB test results within
C Si Mn P S Cr Mo
the range of 40 to 110°F (4 to 43°C) under 0.021 MPa
H2S conditions.7 Sponseller, et al., evaluated SSC 0.25 0.28 0.47 0.006 0.001 1.04 0.72
susceptibility in mildly sour conditions at both room
temperature and 4°C systematically.9 These reports
indicated that SSC susceptibility clearly increased TABLE 2
with decreasing testing temperature. The remarkable Mechanical Properties of Material
influence is different from the conventional knowl- Yield Strength Tensile Strength HRC
edge about the influence of testing temperature on
805 MPa 876 MPa 27.2
the SSC test result.
From this background, a detailed investigation
about the effect of testing temperature on SSC near TABLE 3
room temperature is needed. In this study, effects of Test Solutions
testing temperature on SSC susceptibility were evalu-
NaCl NaAc HAc Initial H2S
ated using a DCB test and a four-point bend (4PB)
(wt%) (wt%) (wt%) pH (MPa)
test in sour environments at several testing tempera-
tures within the range of 4 to 44°C. 4PB test speci- Sol. A 5.0 — 0.5 2.7 0.1
mens included notched and un-notched specimens to Sol. B′ 5.0 0.4 (2.4) 3.5 0.1

604 CORROSION—JUNE 2018

 CORROSION SCIENCE SECTION

saturation with H2S was maintained until the end of the saturated calomel electrode. The test cell of the cathodic
immersion test. side (hydrogen absorption side) was filled with the test
solution.
The hydrogen permeation current was evaluated
Double Cantilever Beam Tests
by long-term hydrogen permeation tests. The test
DCB tests were carried out according to NACE
solutions were solutions A and B′. The testing tem-
TM0177 method D to investigate the effect of testing
peratures were 4, 14, 24, and 34°C, and the test du-
temperature on KISSC values.1 DCB test is a crack
ration was 336 h. The hydrogen diffusion coefficient D
arrest type of fracture mechanics approach, and the
was evaluated by the fitting with the theoretical
SSC susceptibility of steels is evaluated through a
curve using short-term hydrogen permeation tests’
critical stress intensity factor, KISSC. A crack propagates
results.14 As-tempered material and cold-worked
from the root of the beams and stops due to a decrease
material were used to evaluate the effects of testing
in applied KI. From the crack length and applied stress
temperature and plastic deformation on the hydrogen
between the two beams by the wedge after the im-
diffusion coefficient in order to estimate the hydrogen
mersion test, KISSC value can be calculated. A high KISSC
concentration at both un-deformed and deformed
value indicates that the steel has high SSC resistance
areas. Cold working was applied to the material by
under the test conditions, while a low KISSC value
cold rolling to decrease the thickness by 40%. The test
indicates that the steel has low SSC resistance. In this
solution was solution A. The testing temperatures
study, standard-size DCB test specimens with an arm
were 4, 24, and 44°C. The surface hydrogen concen-
displacement of 0.51 mm were used. The test solu-
tration of material C can be estimated from the hy-
tions were both solutions A and B′. The testing tem-
drogen permeation current J and the hydrogen diffusion
peratures were 4, 14, 24, and 34°C. The test duration
coefficient D using the following equation;
was 336 h. The number of the examinations of each
condition is six. After the immersion, corrosion rates
106 × JL
were measured by a weight loss method. The number of C= (1)
each condition is four in six DCB specimens. FDSd

where F is Faraday constant (96500 C/mol), S is

Four-point Bend Test permeation area, and d is the density of iron
4PB tests were carried out according to EFC 16 (7.86 g/cm3).
and ASTM G39 in solution A to investigate the effects of
testing temperature and stress concentration factor α Absorbed Hydrogen Measurement
on SSC susceptibility.2,12 The test specimen geometry The effects of testing temperature and plastic
was a plate with 2 × 10 × 75 mm3 volume. Notched deformation on the hydrogen concentration were in-
and un-notched specimens were used. Several stress vestigated. The plastic deformation was applied by a
concentration factor conditions were prepared for uni-axial tensile method using the round-bar tensile
notched specimens by changing the notch radius. specimens; 5, 10, and 15% of strain were applied to
The stress concentration factor ranged from 1.0 (un- the specimens because the total elongation of the
notched) to 3.0. Threshold stresses of SSC (the lowest material was below 20%. After applying the plastic
stress a specimen fails) were evaluated by changing deformation, the specimens were immersed in solution
the applied stress on the specimen. The actual principal A with 0.1 MPa H2S without applied stress. The
stress at the bottom of the notch can increase up to testing temperatures were 4 and 24°C. After the im-
approximately three times of yield stress, 3σYS, because mersion tests, the absorbed hydrogen concentrations
of the influence of plastic constraint. The testing in the specimens were measured by thermal desorption
temperatures were 4, 24, and 44 °C. The test duration spectrometry (TDS) at a heating rate of 10°C/min and
was 336 h. the hydrogen concentrations were derived by integrating
the hydrogen evolution rate from room temperature
Hydrogen Permeation Test to 600°C.
Hydrogen permeation tests were carried out to
investigate the behaviors of hydrogen’s entry and dif- RESULTS
fusion in the material. The Devanathan-Stachurski
double cell method was used for the hydrogen perme- H2S Concentration in Aqueous Solution
ation current measurement, consisting of a hydrogen The H2S concentrations in solution A saturated
detecting side and a hydrogen absorption side.13 The with 0.1 MPa H2S at several temperatures derived by an
hydrogen permeation test specimen geometry was a iodometric titration method are plotted in Figure 1.
disk with 1.5 mm thickness and 19.6 mm2 permeation The effect of testing temperature on H2S concentration
area. The anodic side (hydrogen detecting side) was in an aqueous solution was calculated using Factsage
plated by Ni. The test cell of the anodic side was filled thermo-chemical software and databases, as shown in
with 1N NaOH solution and polarized at 0 V vs. Figure 1.15-16 The experimental results agreed well

CORROSION—Vol. 74, No. 6 605

CORROSION SCIENCE SECTION

8,000 40
Calculated value
Measurement result
H2S Concentration (ppm)

6,000 30

KISSC (MPa√m)
4,000 20
Solution A (pH 2.7)

Solution B (pH 3.5)

2,000 10

0 0
0 10 20 30 40 50 0 10 20 30 40
Testing Temperature (°C) Testing Temperature (°C)
FIGURE 1. Effect of temperature on H2S concentration in aqueous FIGURE 2. Effect of testing temperature on KISSC value (DCB test,
solution (0.1 MPa H2S). 0.1 MPa H2S).

with the theoretical results. H2S concentration in the 3

Solution A (pH 2.7)
aqueous solution contacted with H2S gas phase in-
creased with decreasing temperature under the same Solution B (pH 3.5)
Corrosion Rate (mm/y)

partial pressure of H2S. Since H2S in an aqueous

solution works as a poisonous catalyst to promote the 2
entry of hydrogen in the material, the dependence of
temperature on H2S concentration is an important
factor for understanding the effect of testing tem-
perature on SSC susceptibility. In the case of the 1
low temperature side, there was much H2S concen-
tration to contribute toward corrosion reaction and
hydrogen entry.
0
0 10 20 30 40
Effect of Testing Temperature on SSC
Testing Temperature (°C)
Susceptibility
The effects of testing temperature on KISSC values FIGURE 3. Effect of testing temperature on corrosion rate (DCB test
specimen, 0.1 MPa H2S, 336 h, weight loss method).
obtained from DCB tests in solutions A and B′ are
shown in Figure 2. The results clearly showed that a
decrease in testing temperature decreased the KISSC
values. The magnitude of the effect was about the corrosion product produced at 4°C and the right
p picture shows that produced at 24°C. At 4°C, a thin
−0.5 MPa m per °C. The KISSC values derived in
solution A were almost equivalent to those in solution and sparse corrosion product (15 μm thick) was ob-
B′ without reference to the test solution’s composition served using a scanning electron microscope (SEM).
and initial pH. Thus, the influence of buffer capacity On the other hand, a thick and dense corrosion product
on KISSC values was considered to be minor. (60 μm thick) was observed at 24°C. The thickness of
The effect of testing temperature on corrosion the corrosion products corresponded to the results of
rates derived by the weight loss method using the DCB corrosion rate.
test specimens is shown in Figure 3. The corrosion The effect of testing temperature on the threshold
rates depended on the testing temperature. Corrosion stresses obtained from 4PB test results of notched and
rates at lower temperature were lower than those at un-notched test specimens are shown in Figure 5 for
higher temperature. The difference between the two test each testing temperature. SSC threshold stress was
solutions was also observed. Corrosion rate in solu- calculated as the product of the applied stress and the
tion B′ was higher than that in solution A, although the stress concentration factor. At 44°C, no crack was
initial pH of solution B′ was higher than that of observed even under the high applied stress and high
solution A. It seems that the increase in the amount stress concentration. At 24°C, several cracks were
of HAc and NaAc increased the corrosion rate of the observed in the high applied stress condition. At 4°C,
material. Corrosion products on the specimens in so- high crack susceptibility was observed compared to
lution A are shown in Figure 4. The left picture shows those at 24 and 44°C.

606 CORROSION—JUNE 2018

 CORROSION SCIENCE SECTION

Hydrogen Permeation Current (mA)

Corrosion 34°C
product 24°C
14°C
4°C

0
0 48 96 144 192 240 288 336
Time (h)
FIGURE 6. Effect of testing temperature on hydrogen permeation
Base current (solution A, 0.1 MPa H2S).
metal
10 μm 4°C 24°C

FIGURE 4. Corrosion products on the DCB specimens produced at 2

Hydrogen Permeation Current (mA)

4°C (left) and 24°C (right) environment (solution A, 0.1 MPa H2S, 34°C
336 h, SEM). 24°C
14°C
4°C
3σ ys
1
44°C
SSC Threshold Stress (MPa)

2,000 24°C
4°C
1,500

0
1,000 0 48 96 144 192 240 288 336
Time (h)

500 FIGURE 7. Effect of testing temperature on hydrogen permeation

current (solution B′, 0.1 MPa H2S).

0
1.0 1.5 2.0 2.5 3.0
Stress Concentration Factor permeation current corresponded to the results of
FIGURE 5. Effect of testing temperature on SSC threshold stress corrosion rate for both test solutions.
(4PB test, solution A, 0.1 MPa H2S, 3σYS is 2,415 MPa). The effects of cold working and testing temper-
ature on the hydrogen diffusion coefficient D are shown
in Figure 8. The results clearly show that the hy-
drogen diffusion coefficients decreased with a decrease
Effects of Testing Temperature and Cold Working in testing temperature. In addition, these coefficients
on Hydrogen Absorption were remarkably decreased by cold working.
Effects of testing temperature on the hydrogen Surface hydrogen concentrations in the steels
permeation current, J, derived by the hydrogen per- can be calculated from the hydrogen permeation cur-
meation test are shown in Figures 6 and 7. The rent and the hydrogen diffusion coefficient using
hydrogen permeation currents just after H2S saturation Equation (1) under the supposition that the hydrogen
at 34°C were the largest at all testing temperatures permeation rate is not affected by cold working. Since
and decreased with a decrease in testing temperature. the hydrogen absorption behavior depends on the
The hydrogen permeation currents gradually decayed surface reaction on steel, the influence of cold
during the test, which was observed at higher tem- working on the hydrogen permeation rate is considered
peratures. We consider that the decay of hydrogen to be small. The calculation results are shown in
permeation current was due to the production of cor- Figure 9. An average of hydrogen permeation rates
rosion product layer. The behavior of the hydrogen during the test was used for the calculation. In the

CORROSION—Vol. 74, No. 6 607

CORROSION SCIENCE SECTION

4
Hydrogen Diffusion Coeffecient D (cm2/s)

1.0×10–5

Relative Hydrogen Concentration

As tempered 4°C

40% cold work 24°C

3

1.0×10–6

1.0×10–7
1

1.0×10–8 0
3.0 3.2 3.4 3.6 3.8 0 5 10 15 20
1,000/T (K–1) Applied Strain (%)

FIGURE 8. Effects of cold working and testing temperature on FIGURE 10. Effects of tensile strain and testing temperature on the
hydrogen diffusion coefficient (hydrogen permeation test, solution A, absorbed hydrogen concentration.
0.1 MPa H2S).

on un-strained specimens. The test results showed that

250 tensile strain of over 10% increased the absorbed
Hydrogen Concentration (ppm)

As tempered hydrogen concentration, especially at lower tempera-

40% cold work tures. The increase in absorbed hydrogen concen-
200
tration by tensile strain can be attributed to the increase
in lattice defects by deformation.
150

DISCUSSION
100
From the DCB test results, KISSC values strongly
depended on testing temperature, as shown in Figure 2.
50 The decrease in testing temperatures decreased KISSC
values in both solution A and solution B′. From 4PB
0 tests, the decrease in testing temperatures decreased
0 10 20 30 40 SSC threshold stress, as shown in Figure 5. The effects
Testing Temperature (°C) of testing temperature on SSC susceptibility will be
discussed from the prospective of hydrogen entry, ab-
FIGURE 9. Calculated surface hydrogen concentration using both
the hydrogen permeation rate and hydrogen diffusion coefficient sorption, and accumulation to the crack initiation
(solid mark: solution A, open mark: solution B′). and propagation area.

Effect of Testing Temperature on Hydrogen

Absorption
case of as-tempered materials, the hydrogen concen- The influences of environmental temperature on
trations were almost equivalent regardless of the corrosion reaction, hydrogen entry, and hydrogen dif-
testing temperature and test solution; the hydrogen fusion coefficient are concluded as follows. Corrosion
concentration was about 10 ppm. On the other hand, rate in sour environment decreased with decreasing
in the case of 40% cold-worked materials, very high testing temperature, as shown in Figure 3. The hy-
hydrogen concentration was estimated compared to drogen permeation current also decreased with de-
the as-tempered materials. The hydrogen concentration creasing testing temperature, as shown in Figures 6
was 60 to 200 ppm, and ranged from 6 to 20 times for and 7. The hydrogen diffusion coefficient decreased with
as-tempered material. It was strongly affected by testing decreasing testing temperature, especially for cold-
temperature and remarkably increased with a de- worked material, as shown in Figure 8. From these
crease in testing temperature. influences, the effect of temperature on hydrogen
The effect of tensile strain on the absorbed absorption will be discussed.
hydrogen concentration derived using round-bar The absorbed hydrogen concentration is the
specimens is shown in Figure 10. In this figure, the most important factor for understanding the SSC
absorbed hydrogen concentration is plotted as a susceptibility of steel. In general, the increase in
function of strain relatively to its value measured hydrogen concentration increases SSC susceptibility

608 CORROSION—JUNE 2018

 CORROSION SCIENCE SECTION

by comparison in the same steel. As described in were observed. This could not be explained by the
Equation (1), hydrogen concentration in steel can be average hydrogen concentration in the material.
estimated from the hydrogen permeation test result.
While hydrogen concentration is proportional to hy- Effect of Plastic Deformation on Hydrogen
drogen permeation rates, it is inversely proportional
Accumulation
to the hydrogen diffusion coefficient. When the hydro-
In the cases of DCB tests, the local hydrogen
gen diffusion of steel is fast, the hydrogen that entered
accumulation at the stress concentration area should
the steel rapidly runs away from the steel. On the other
be discussed to explain the effect of testing temper-
hand, when the hydrogen diffusion is slow, the hy-
ature on SSC susceptibility, because SSC occurrence
drogen stays for a long time in the steel. The reciprocal of
is dominated by the local stress and local hydrogen
D indicates the index of hydrogen trapping ability W,
concentration at the stress concentration area. How-
as shown in Equation (2):
ever, the local hydrogen concentration of the test
specimen cannot be measured using existing hydrogen
1
W∝ : (2) measurement methods because of the high mobility
D of hydrogen in steel even at room temperature. Thus,
The increase in testing temperature increased the it has to be estimated or calculated as follows.
corrosion rates, as shown in Figure 3. In a sour corro- The increase in hydrogen concentration associated
sion environment, the major cathodic reaction that with the applied tensile stress has been reported.17-19
accompanies the corrosion reaction is the hydrogen The hydrogen concentration under tensile stress
generation reaction on steel because of both low pH CS can be obtained with the following equation:
and lack of dissolved oxygen. The amount of hydrogen  
generated on the steel depends on the corrosion rate. Δσ · VH
CS = C0 exp , (3)
Although a part of the hydrogen released from the 3RT
surface of the steel as hydrogen gas, much hydrogen
was absorbed into the steel due to the poisonous cat- where C0 is the hydrogen concentration in bulk ma-
alyst effect of H2S. Therefore, the temperature de- terial, Δσ is applied stress, VH is the molar volume
pendence of the hydrogen permeation current shown in of hydrogen (2 × 10−6 m3/mol), R is the gas constant
Figures 6 and 7 is comprehensible. The behavior of (= 8.31 J/K·mol), and T is the absolute temperature.
the hydrogen permeation current corresponds to the Using this equation, the effect of temperature on the
corrosion rate in sour corrosion environments. It is hydrogen concentration under tensile stress was esti-
assumed that a decrement of hydrogen permeation rate mated, as shown in Figure 11. Although the relative
during the test at higher temperature is caused by hydrogen concentration increases with decrease in
decreasing corrosion rate, because a thick and dense temperature, the magnitude of the effect is very small.
corrosion product forms on the specimen surface, as In addition, the maximum stress at crack tip area is
shown in Figure 4. It is thought that the decay of considered to be increased with a decrease in testing
hydrogen entry during the test influences the SSC temperature because the yield stress increases at lower
test result; however, it is not clear correlation between temperature. However, the effect of testing tempera-
hydrogen entry rate and crack propagation. At lower ture on local hydrogen concentration seems to be
temperature, the corrosion rate and hydrogen perme-
ation current were lower than those at higher tem-
perature. However, the hydrogen permeation current 1.6
Relative Hydrogen Concentration

was sufficiently high to cause SSC even at 4°C in the 4°C

case of solutions A and B′ saturated with 0.1 MPa of 1.5 24°C
H2S. The high H2S concentration at low temperature 44°C
1.4
shown in Figure 1 can be attributed to the hydrogen
entry. 1.3
The hydrogen diffusion coefficient of the as-
tempered material at low temperature was smaller than 1.2
that at high temperature, as shown in Figure 8. The
effect of testing temperature on hydrogen diffusion co- 1.1
efficient is opposite to that on hydrogen entry from the
1.0
viewpoint of hydrogen concentration, as shown in
Equation (2). From these contributions, the estimated 0.9
hydrogen concentration was almost equivalent regard- 0 500 1,000 1,500
less of the testing temperature, as shown in Figure 9. Applied Stress (MPa)
In the cases of DCB and 4PB tests, however, remarkable FIGURE 11. Effect of tensile stress on the relative hydrogen
effects of testing temperature on SSC susceptibility concentration.

CORROSION—Vol. 74, No. 6 609

CORROSION SCIENCE SECTION

insufficient to explain the effect of testing tempera-

Relative Hydrogen Trapping Ability (at 24°C)

ture on SSC susceptibility. 4
Cold working with plastic deformation increased 4°C Vacancy
the surface hydrogen concentration, as shown in 24°C
Figure 9. The increase in hydrogen concentration is 3 44°C
remarkable, especially at lower temperature. In case
of tensile plastic deformation using the round-bar Dislocation
specimens, the same tendency to increase the absorbed 2
hydrogen concentration derived by TDS was ob-
α-iron
served, as shown in Figure 10. The reason why plastic
deformation leads to an increase in absorbed hy- 1
drogen concentration is that plastic deformation
increases lattice defects, which could be stable hy-
drogen trapping sites. The difference in magnification 0
of the hydrogen increment with plastic deformation 0 10 20 30 40 50
between the two evaluation methods is considered to Bonding Energy HB (kJ)
be caused by the difference in the volume rate of FIGURE 12. Effect of lattice defect on local hydrogen trapping ability.
deformation. In the case of a tensile test specimen, the
plastic deformation is localized.
The temperature dependence of the increase in
local hydrogen accumulation should occur according to
absorbed hydrogen concentration with plastic defor-
the environmental temperature, as shown in
mation can be explained by the hydrogen trapping
Figure 12.
behavior of lattice defects. When lattice defects (dislo-
It is assumed that plastic deformation is intro-
cations, vacancies) are created by plastic deforma-
duced at the stress concentration area such as the crack
tions, these defects play roles of hydrogen trapping
tip area in the DCB specimen. Many lattice defects are
sites. The hydrogen diffusion coefficient D depends on
produced in the local area, which can trap large
the bonding energy between lattice defects and hydro-
amounts of hydrogen, especially at low temperature,
gen HB. D is derived from Oriani’s equation (4), where
as illustrated in Figure 13. This can explain why a
D0 is the hydrogen diffusion coefficient of the lattice site
(4.2 × 10−8 m2/s: BCC Fe-H), Nx is the number of trap decrease in testing temperature leads to a decrease in
sites per unit volume, and NL is the number of lattice KISSC values. This hypothesis is supported by the effect
sites per unit volume. Nx/NL means lattice defect of cold work on the temperature dependence of hy-
density:20 drogen diffusion coefficient, as shown in Figure 8. The
activation energy of hydrogen diffusion Ea can be
estimated from the slope using the Arrhenius equation.
D0
D= : (4) The activation energy of the cold work specimen
1 þ expðHB =RTÞðNx =NL Þ
(41 kJ/mol) was higher than that of the un-strained
When the lattice defect density is sufficiently specimen (22 kJ/mol). These activation energies
high, Equation (4) can be rewritten as follows: show that cold working increases the degree of hydrogen
accumulation at low temperatures due to increasing
   lattice defects.
−HB NL
D ≈ D0 exp : (5)
RT Nx
Stress σ Stress σ

From Equation (5), the effect of temperature and

bonding energy of lattice defects on the hydrogen
trapping ability W is calculated using their HB (Pure - H
4°C H
24°C
α iron: 3.8 kJ/mol, dislocation: 25 kJ/mol, vacancy
46 kJ/mol), as shown in Figure 12.20-21 In this figure,
the hydrogen trapping abilities at 4 and 44°C are plotted HHHHH H H HH H
as a function of bonding energy relative to its value HHHH HH
calculated at 24°C. From the results, the hydrogen HHH HH HH H
H
trapping ability increases with decreasing tempera-
ture at the local area with high lattice defect density. In H H
wet H2S environments, the local hydrogen concen-
tration should be proportional to the hydrogen trapping
ability because it can be supposed that there is suf-
FIGURE 13. Schematic of hydrogen accumulation at the crack tip
ficient hydrogen absorption to fill the new hydrogen trap
area or high-stress concentration area.
sites produced by plastic deformation. Therefore, the

610 CORROSION—JUNE 2018

 CORROSION SCIENCE SECTION

In the case of the 4PB tests, a plastic deformed the amount of hydrogen absorption is much lower
area did not exist definitely because the SSC threshold than that in sour environments. In this situation, the
stress was under the yield stress, as shown in Figure 5. hydrogen generation and entry are the dominant
However, a local plastic deformed area on the 4PB factors that cause hydrogen embrittlement. The rela-
specimen was considered to be generated by pittings tively high temperature at which sufficient corrosion
and fissures with corrosion. Consequently, the high occurs is the most severe temperature for hydrogen
hydrogen accumulation at the local area at lower embrittlement.
temperature can be expected.
CONCLUSION
Effect of Testing Temperature for HE Test Method
For evaluation methods with high stress con- Effects of testing temperature on SSC suscepti-
centration, the increment in SSC susceptibility at lower bilities of low-alloy steel were investigated by DCB and
temperature could be explained by considering the 4PB tests. Then, these effects were discussed from the
local hydrogen accumulation at the plastic deformed perspectives of hydrogen entry, absorption, and accu-
area. In other words, an occurrence of high SSC mulation to the crack initiation and propagation area.
susceptibility at low temperature needs both plastic The results are as follows.
deformation at the high stress area and sufficient v KISSC values derived by the DCB test strongly
hydrogen entry to increase the hydrogen concentration depended on testing temperature. The decrease in
at the local area. testing temperature decreased the KISSC values. The
p
It is considered that NACE TM0177 method D magnitude of the effect was about −0.5 MPa m per °C.
(DCB test) is the most sensitive to the testing temper- v The decrease in testing temperature decreased the
ature among all established SSC evaluation methods. threshold stresses of SSC derived by 4PB tests.
When the steel is used in a situation such as a DCB test, v The hydrogen concentration at the plastic de-
a noticeable deterioration of SSC resistance takes formed area increased remarkably with decreasing
place at low temperature. Therefore, a prior material testing temperature.
evaluation reproducing both environmental temper- v For explaining the temperature dependence on
ature and actual stress condition is needed. Otherwise, SSC susceptibility, a mechanism in which the degree
the usage with stress concentration and plastic de- of hydrogen accumulation at the local plastic defor-
formation has to be avoided in a sour environment, mation areas can be higher at lower temperature than
especially at low temperatures. The magnitude of the that at higher temperature due to the lattice defects
effect of testing temperature on SSC susceptibility produced is proposed.
depends on the evaluation method. v The magnitude of the effect of testing temperature
The duration of SSC test is also important when on SSC susceptibility depends on the evaluation
the effect of testing temperature on SSC susceptibility is method. The material evaluation reproducing both
evaluated. The hydrogen diffusion coefficient is strong the actual environmental temperature and the actual
in temperature dependence, as shown in Figure 8. At stress condition is important.
low temperature, the hydrogen diffusion coefficient is
small. Therefore, longer time is necessary for the ac- ACKNOWLEDGMENT
cumulation of hydrogen to the local plastic deformed
area at lower temperature. SSC test results under fixed The authors thank Nippon Steel & Sumitomo Metal
test duration may lead to incorrect conclusions. In Corporation for allowing publication of this study. The
this study, it was confirmed that the test duration for assistance and discussion of co-workers are also
DCB and 4PB tests were long enough even at 4°C from gratefully acknowledged.
the specimen size and the diffusion coefficient. The
effect of temperature on hydrogen accumulation
discussed in this paper is an essential phenomenon REFERENCES
and independent of both material and environment.
1. “Standard Test Method, Laboratory Testing of Metals for Resis-
There is a possibility that the susceptibility to hydrogen tance to Sulfide Stress Cracking and Stress Corrosion Cracking in
embrittlement increases at lower temperature when H2S Environments,” ANSI/NACE TM0177, (Houston, TX: NACE
sufficient hydrogen is absorbed in steels. In a sour International, 1996, 2005, and 2016).
2. “Guidelines on Materials Requirements for Carbon and Low Alloy
environment with liquid phase water, a large quantity Steels for H2S-containing Environments in Oil and Gas Produc-
of hydrogen is absorbed by the corrosion reaction on low tion,” European Federation of Corrosion Publications, No. 16, 3rd
Edition (2009).
alloy steel even at 4°C. The SSC susceptibility 3. “Petroleum, Petrochemical and Natural Gas Industries - Materials
increases with decreasing temperature. On the other for Use in H2S-containing Environments in Oil and Gas Production -
hand, when there is no liquid phase water, hydrogen Part 2: Cracking-resistant Carbon and Low Alloy Steels, and the
Use of Cast Irons,” ANSI/NACE MR0175/ISO 15156-2:2015,
absorption due to corrosion reaction is impossible. SSC (ANSI/NACE/ISO, 2015).
does not occur at below the freezing point of water. 4. H.E. Townsend Jr., Corrosion 28 (1972): p. 39-46.
Additionally, in the atmospheric corrosion environment, 5. K.A. Kvenvolden, Rev. Geophys. 31 (1993): p. 173-187.

CORROSION—Vol. 74, No. 6 611

CORROSION SCIENCE SECTION

6. G.C. Guntz, C.P. Linne, F. Puissochet, “DCB test—A Review of 12. “ASTM G39—Standard Practice for Preparation and Use of Bent-
Critical Parameters,” Proceedings of Corrosion/99, paper no. 606 Beam Stress-Corrosion Test Specimens,” (ASTM, 2016).
(Houston, TX: NACE, 1999). 13. M.A.V. Devanathan, Z. Stachurski, Proc. R. Soc. A 270 (1962):
7. B.P. Miglin, L.A. Skogsberg, W.D. Grimes, “Sulfide Stress Cracking p. 90.
of Low Alloy Steel - Uniaxial Tension and DCB Tests in Modified 14. J. McBreen, L. Nanis, W. Beck, J. Electrochem. Soc. 113 (1966):
NACE Environment,” Proceedings of Corrosion/2005, paper no. p. 1218-1222.
05086 (Houston, TX: NACE, 2005). 15. C.W. Bale, P. Chartrand, S.A. Degterov, G. Eriksson, K. Hack, R.
8. K. Kobayashi, T. Omura, M. Ueda, K. Nakamura, “Effect of Testing Ben Mahfoud, J. Melancon, A.D. Pelton, S. Peterson, Calphad J. 26
Temperature on SSC Properties of Low Alloy Steel,” Proceedings of (2002): p. 189-228.
Corrosion/2006, paper no. 06127 (Houston, TX: NACE, 2006). 16. C.W. Bale, E. Belisle, P. Chartrand, S.A. Decterov, G. Eriksson, K.
9. D.L. Sponseller, T.E. Sponseller, B.E. Urband, “Sulfide Fracture Hack, I.H. Jung, Y.B. Kang, J. Melancon, A.D. Pelton, C. Robelin, S.
Toughness of Low-alloy Steel OCTG in Mild Conditions at Room Peterson, et al., Calphad J. 33 (2009): p. 295-311.
Temperature and 40°F,” Proceedings of Corrosion/2008, paper no. 17. J.C.M. Li, R.A. Oriani, L.S. Darken, Zeitschrift fur Physikalische
08107 (Houston, TX: NACE, 2008). Chemie Neue Folge 49 (1966): p. 271-290.
10. M.J. Cancio, T.E. Perez, J.R. Collet-Lacoste, “A Mechanistic 18. H.A. Wriedt, R.A. Oriani, Acta Metall. 18 (1970): p. 753-760.
Approach to Sulfide Stress Cracking Resistance of High Strength 19. T. Doshida, H. Suzuki, K. Takai, N. Oshima, T. Hirade, ISIJ Int. 52
OCTG Steels,” Proceedings of Corrosion/2009, paper no. 09098 (2012): p. 198-207.
(Houston, TX: NACE, 2009). 20. R.A. Oriani, Acta Metall. 18 (1970): 147-157.
11. T. Omura, K. Kobayashi, M. Ueda, “SSC Resistance of High 21. S.M. Myers, S.T. Picraux, R.E. Stoltz, J. Appl. Phys. 50 (1979):
Strength Low Alloy Steel OCTG in High Pressure H2S Environ- p. 5710.
ments,” Proceedings of Corrosion/2009, paper no. 09102
(Houston, TX: NACE, 2009).

612 CORROSION—JUNE 2018