See discussions, stats, and author profiles for this publication at: [Link]

net/publication/273755740

HIC resistance of heritage pipelines exposed to mildly sour environments

Conference Paper · March 2013

CITATIONS READS
5 514

7 authors, including:

Gareth Hinds
National Physical Laboratory
177 PUBLICATIONS 5,708 CITATIONS

SEE PROFILE

All content following this page was uploaded by Gareth Hinds on 17 July 2015.

The user has requested enhancement of the downloaded file.

 HIC Resistance of Heritage Pipelines Exposed to Mildly Sour Environments

Viviane C.M. Smith John W. Martin

BP Exploration & Production BP Exploration & Production
Chertsey Street, Sunbury-on-Thames, Chertsey Street, Sunbury-on-Thames,
Surrey, TW16 7NL Surrey, TW16 7NL
UK UK

Gareth Hinds Natalie Street

National Physical Laboratory National Physical Laboratory
Hampton Road, Teddington, Hampton Road, Teddington,
Middlesex, TW11 0LW Middlesex, TW11 0LW
UK UK

Christoph Bosch Thomas Haase

Salzgitter Mannesmann Forschung Salzgitter Mannesmann Forschung
Ehinger Straße 200, 47259 Ehinger Straße 200, 47259
Duisburg Duisburg
Germany Germany
Stefan Jäger
Salzgitter Mannesmann Forschung
Ehinger Straße 200, 47259
Duisburg
Germany

ABSTRACT

Carbon and low alloy steel notionally sweet pipelines could be in danger of sudden and/or severe
cracking and, consequently, loss of containment if H2S is introduced into the media, even for a short
time and/or at very low partial pressures. For ‘mildly sour’ conditions i.e. Region 0 as defined in
NACE MR0175/ISO 15156, the main threat is hydrogen induced cracking (HIC) and this needs to be
considered even for trace amounts of H2S. The effect of long-term exposure at such low levels of H2S is
not well defined and the window of safe operation for HIC resistance is yet to be established. Recent
work carried out by the European Pipeline Research Group suggested that there is a threshold below
which HIC may not be a credible threat. This work however used modern seam-welded linepipe, which
is inherently cleaner and thus potentially less susceptible to HIC than heritage materials manufactured
decades ago. The present paper is aimed at defining a threshold and assessing the applicability of the
standard test methodology for the HIC resistance of heritage pipelines (i.e. manufactured decades
ago).

Key words: Hydrogen Sulphide, Hydrogen Induced Cracking, ‘Heritage’ Materials

 INTRODUCTION

Carbon and low alloy steel sweet pipelines could be in danger of sudden and/or severe cracking and,
consequently, loss of containment if hydrogen sulphide (H2S) is introduced into the media, even for a
short time and/or at very low partial pressures of H2S. The main integrity threats associated with low
partial pressures of H2S are (i) Sulphide Stress Cracking (SSC) and (ii) Hydrogen Induced Cracking
(HIC). The SSC threat is fairly well understood and NACE MR0175/ISO 15156,1 provides guidance to
avoid this threat. However, the effect of low levels of H2S on the HIC resistance is not well defined.
NACE MR0175/ISO 15156 requires that the user considers HIC resistance even for cases where trace
amounts of H2S are present. This is often interpreted as the presence of a measurable amount of H2S
(normally around 1 ppm by volume in the associated gas phase) and includes Region 0 defined for
SSC in NACE MR0175/ISO 15156.

This poses a dilemma for mildly sour conditions (i.e., above 1ppm H2S, but still within Region 0) since
there are no well-defined or broadly accepted means of demonstrating resistance to HIC. The standard
laboratory test method for qualification of HIC resistance, NACE TM0284,2 can be used for evaluating
the HIC resistance of seam-welded linepipe but the test method is believed to be overly conservative
for the assessment of material in mildly sour conditions. For example, fully immersed specimens are
employed, resulting in hydrogen charging conditions that are severe and not representative of those in
a pipeline, which is exposed to H2S-containing fluids only on one side (internally). This is compounded
by a ‘no cracking’ acceptance criteria whereby permitting a small amount of cracking is perhaps more
reasonable given the aggressive exposure conditions. Linepipe that has been manufactured for sweet
service may not pass this screening test, especially old lines, which may employ materials that may not
have the quality of more recently manufactured material. A recently developed alternative is to perform
a full-scale “fitness-for-service” test using a "full ring" test section of linepipe exposed internally to
simulated field sour fluids3. This method can, however, be ill suited for all applications since there are
practical limits to the number of such tests that can be undertaken. A smaller scale “screening” test
method is desirable as an intermediate to performing full scale testing and, for the full scale “fitness-for-
purpose” test, better definition of the required exposure duration is required to allow sufficient time for
any cracking to develop and prevent non-conservative results.

The HIC resistance is influenced by the steel composition and the manufacturing route employed. The
level of sulfur in the steel is particularly important; NACE MR0175/ISO 15156 recommends a maximum
level of 0.003 % mass fraction for seam-welded products, but this limit is merely indicative and there
have been examples of field failures occurring in ‘clean’ steel4-6. Recent work carried out by the
European Pipeline Research Group (EPRG)7,8 suggests that there is a threshold partial pressure of H2S
below which HIC may not be a credible threat and a preliminary limit between 0.1 and 1 kPa (0.001 and
0.01 bar or 0.0145 and 0.145 psi) of H2S has been proposed. This work only used recently
manufactured seam-welded linepipe, which is normally inherently cleaner than ‘heritage’ steels
manufactured decades ago and therefore does not account for the potential effects of ‘old’ versus ‘new’
steel making routes.

In the present paper, work presents the results of HIC full-scale ring test of a approximately 30 year old
linepipe under several sour environments, in absence of applied load. Tests were monitored using
acoustic emission (AE), ultrasonic inspection (UT) and hydrogen probe, and were subsequently
sectioned to validate the presence, or otherwise, of cracking. Test durations were for up to 6 months.
Results are compared with those from modified NACE TM0284 tests of the same material exposed for
similar durations.

The factors influencing the morphology of HIC (i.e. ‘stepping’ cracks) has not been considered in the
present study but could have been assisted by residual or applied stresses, as well as chemical
composition.
 EXPERIMENTAL PROCEDURE

Materials

An approximately 30 year old 915 mm (36 in) diameter API 5L X65 seam-welded pipeline steel was
employed in this work. No girth weld was included in the tested pipe sections. The chemistry of the
material was determined by optical emission spectroscopy (OES). Conventional metallographic
methods were used to examine the through thickness microstructure of the parent materials and
weldment and the cleanliness of the steel.

Small Scale Tests

Testing.

Small-scale testing was carried out in accordance with NACE TM0284 in simulated service conditions.
Triplicate specimens were extracted at the seam weld and 90° from the seam weld. Testing at 180°
from the seam weld was not possible as only one half of the pipe was available. The specimens were
prepared as per NACE TM0284 and immersed in a 5 wt% sodium chloride test solution. The solution
was purged with nitrogen at 100 ml/min for 1 hour per litre of solution prior to introducing the test gas
(i.e. H2S/CO2 mixture), the pH-value was then adjusted to pH 5.2 using sodium bicarbonate. Test
durations of 1 and 3 months were used. A temperature of 25°C ± 3°C (77°F ± 5°F) was maintained
throughout the testing period.

Temperature and gas flow were monitored twice per working day. The pH was initially measured once
per working day, then fortnightly once the approximate rate of pH change had been established. The
pH was adjusted to the target value ± 0.2 pH units using concentrated HCl as required.

At the end of the test, the test solution was purged with nitrogen prior to recovering the specimens,
which were then 100% visually and volumetrically inspected (using ultrasonic technique). When
present, crack indications were evaluated by sectioning as per NACE TM0284. Cracks have been
reported as a % crack surface ratio (% CSR), crack length ratio (% CLR), and crack thickness ratio (%
CTR) values. Any pits, blisters and/or cracks were reported.

Non-Destructive Examination.

Scanning acoustic microscopy (SAM) was used to carry out the volumetric inspection, after calibration.
This technique is a non-contact, non-destructive, ultrasonic interrogation method used for determining
elastic properties or for detecting defects in materials. Partial reflection and transmission of the
ultrasound occurs at any boundary with a different acoustic impedance i.e. flaws or material changes.
The largest amplitude reflections invariably occur at gas interfaces due to the large acoustic mismatch
between gas and any solid or liquid medium. As a result, the method is ideal for detecting laminar
cracks in materials.

In this work a Sonix UHR† 2001 scanning acoustic microscope (SAM) was used to scan a number of
metal samples to determine the development of internal damage after environmental exposure. A 15
MHz focussed transducer was used to produce a high frequency ultrasound beam which was
transmitted via water immersion into the sample and also to detect the signals reflected back towards it.
This signal is displayed as an A-scan, defined as the change in the returned signal amplitude with time.
Signal gates are used to collect information at desired regions of the A-scan. The gate is positioned
over a region of the A-scan where a signal of interest is expected and the largest amplitude signal
which breaks the gate threshold is recorded. In this case, the back wall reflection was monitored so that
any intervening defects would cause a noticeable change in the received signal amplitude.
†
Trade name
In addition, some scans were performed where a full A-scan was captured using consecutive gates
effectively “slicing” the sample into layers and enabling the through-thickness position of any signal
sources to be estimated. The data was captured at 50 micron intervals in x and y directions in a raster
scan pattern. Due to the very small back wall signals produced (high attenuation in thick samples), the
system was set-up with maximum energy input, maximum gain and a lower voltage range for
digitisation. All scans were performed over an area of 105 x 26 mm.

Metallographic Sectioning.

Metallographic sectioning of each specimen was carried out as per NACE TM0284.

For comparison, the material was also tested in accordance with NACE TM0284 in solution A. Upon
completion of the test, specimens were examined as described above for testing in simulated service,
i.e., (i) 100% visual inspection; (ii) 100% volumetric inspection using SAM; and (iii) sectioning in
accordance with NACE TM0284.

Full Scale Tests

Pipe Preparation.

1m length full ring sections of pipe with square machined ends (25 mm [1 in] thick and 910 mm [36 in]
diameter) were used for the large scale testing. Any internal surface contamination, paint residues and
any internal corrosion products were removed by grit blasting to Sa 2½ finish using new grit. The pipe
rings were then fitted with gas tight end plates and seals at both ends according to Figure 1 and
vertically positioned on a safety tank. The pipe sections were 100% ultrasonically inspected from the
outer surface to ensure that all indications exceeding the reference levels were recorded prior to
exposure to the corrosive environment.

Figure 1: Assembled Test Ring Mounted with 12 Acoustic Emission Probes and Hydrogen
Probes
Test Procedure.

The assembled test rings were filled with a 5 wt% Sodium chloride solution and purged with nitrogen for
16 hours followed by CO2 purging for 2 hours prior to introducing the test gas (i.e. H2S/CO2 mixture).
The gas test was bubbled through the solution for the duration of the test. The pH-value was then
adjusted to pH 5.2. The test period commenced 24 hours after introduction of the gas mixture. A
temperature of 25°C ± 3°C (77°F ± 5°F) was maintained throughout the testing period. Testing was
carried out for durations of 1, 3, and 6 months.

Temperature, gas flow and pH of the test were monitored twice per working day. The pH was adjusted
to the target value ± 0.2 pH units using 1 N HCl, as required.

The assembled test rings were equipped with a hydrogen probe in order to determine the time-
dependent hydrogen permeation rate. Hydrogen flux was also measured using a handheld probe
throughout the testing period. Crack initiation and growth were continuously monitored by acoustic
emission (AE) and the test ring was ultrasonically inspected on a monthly basis over a period of six
months. Indications exceeding the reference levels were recorded and clearly marked on the sample
OD. Any crack evolution was carefully monitored throughout the test.

Upon completion of the test, the solution was purged by bubbling nitrogen through it to remove any
dissolved test gas before draining. The pipe sections were then purged with nitrogen followed by
removal of the end caps and hydro-jet cleaning of the inner surface.

Non-Destructive Examination.

Ultrasonic Inspection.

The outer surface was inspected ultrasonically prior to testing. Further ultrasonic inspection was carried
out on a monthly basis. Scanning was limited to two-thirds of the pipe surface because of limited
access.

Ultrasonic inspection was carried out using an Echograph 1090†. The pipe body was examined by
normal probes to detect laminar and stepwise indications. The pre-set sensitivity corresponds to a 3
mm disk shaped reflector (DSR). Distance-gain-size (DGS) compensation was used. A 100 mm wide
band along the longitudinal welds was examined from both sides of the welds using 45°and 60° probes
to detect crack-like indications.

After a scan of the complete pipe surface at the end of the test, two 350 mm x 300 mm pipe sections
were selected from the pipe exposed to 10 mbar H2S for additional ultrasonic inspection and
determination of the crack area ratio (CAR). For the purpose, two 400 mm x 400 mm coupons were
flame cut from the pipe sample. The pipe exposed to 0.5 kPa (0.005 bar or 0.07 psi) of H2S underwent
no additional ultrasonic inspection, due to the absence of any indications in the preceding tests, as
supported by acoustic emission results. CAR values were calculated for a 25% screen height threshold.

Acoustic Emission.

Acoustic Emission (AE) is an NDT method employing piezoelectric sensors that operate at a frequency
spectrum of 20 to 500 KHz. It is able to detect elastic transient waves from the rapid release of energy
within solid materials with a wide range of sizes, material forms, geometries, temperatures and
accessibilities. Materials can produce transient elastic waves through crack initiation or propagation.
The most likely mechanism can be determined via evaluation of the time signature of the micro seismic
event.

AE establishes the presence or absence of defects based on the correlation of multiple hits mapping at
the same location. The location can be established through measurement of the ultrasonic time of flight
of each ‘hit’ and comparing it to neighbouring sensor data. Sensors were applied to each pipe segment

†
Trade name
as per ASTM E650 9, in three rows of three sensors at 120 degrees in a staggered formation. Spacing
between sensors was determined such that the distance did not exceed the sensor zone radius.

The test acoustic threshold was set at the data analysis threshold of 40 dB. Circuit verification included
three breaks of 0.5 mm (2H) pencil leads at specific distances from each sensor. Average peak
amplitudes did not vary more than 4 dB from the average of all sensors. The testing environment was
checked so as to eradicate as much extraneous noise as feasibly possible. Calibration was in
compliance with ASTM(1) E750 10, and consistency verification was in compliance with ASTM E2075 11.

Data was monitored remotely on a daily basis; analysis was in accordance with ASTM E1316 12, to
ascertain credible indications whilst disregarding spurious, non-structural sources in the data. Live test
data could only provide preliminary indications of degree of activity. Only interval analysis of the
'duration versus hit' and 'amplitude versus hit' correlations can give more conclusive evidence of true
indications, and their corroboration with failures. After this was completed, the cumulative energy per
channel versus time was used to identify the most active areas, and rank them in order of threat using
the MONPAC letter system. The sensor location data was used in a similar manner, but also to identify
clusters of potential threats as a precursor to ultrasonic testing.

AE signals were collected continuously and results were examined on a fortnightly basis.

Sectioning.

After final ultrasonic inspection, HIC evaluation in accordance with NACE TM0284 was carried out on
micro-sections taken from the two pipe coupons labeled 1A and 1B. The precise sampling positions are
shown in Figure 2.

Figure 2: Location of Pipe Coupons Selected for Metallographic Evaluation

Sections 1A-1, 1A-2 and 1A-4 were investigated in the direction parallel to pipe circumference, while
sections 1A-3 and 1B-1 were investigated perpendicular to pipe circumference, in order to cover as
many cracks as possible in the coupon, with the objective to find interconnections between the different
cracks. Besides determination of CLR, CTR and CSR, all detected cracks were documented via photo-
micrographs at 20x and 200x magnification.

Testing Conditions

The testing conditions for both small and full scale tests in simulated service conditions are summarized
in Table 1 below.

(1)
ASTM International, 100 Barr Harbor Dr., PO Box C700, West Conshohocken, PA, 19428.
 Table 1
Summary of the Testing Conditions
NACE TM0284 Modified Ring Test
Parameter
(Small Scale) (Full Scale)
Partial 0.34 kPa (0.0034 bar or 0.05 psia) -
pressure of 0.5 kPa (0.005 bar or 0.07 psia) 0.5 kPa (0.005 bar or 0.07 psia)
H2S 1 kPa (0.01 bar or 0.15 psia) 1 kPa (0.01 bar or 0.15 psia)
pH 5.2 ± 0.2 5.2 ± 0.2
Temperature 25°C ± 3°C (77°F ± 5°F) 25°C ± 3°C (77°F ± 5°F)
Duration 1 month
6 months
3 months

RESULTS

Microstructure and Chemistry

The chemical composition of the material is given in Table 2.

Table 2
Chemical Composition of the Parent Material
Elements C Si Mn P S Cr Mo Ni Al As B
Wt% 0.10 0.19 1.53 0.011 0.005 0.012 <0.003 0.020 0.039 <0.004 0.0003

Co Cu Nb Pb Sn Ti V W Zr Ca Ce Sb
0.006 0.015 0.034 <0.005 <0.004 0.002 0.030 <0.01 <0.005 <0.0003 <0.002 <0.002

A banded microstructure was observed in the through thickness direction of the parent material
(Figure 3).

a) Parent Material b) Heat Affected Zone

Figure 3: Micrograph Showing Banded Through Thickness Microstructure2

Metallographic examination of non-etched specimens revealed a high density of inclusions (Figure 4a).

(2)
The Test House Ltd, Cambridge
†
Small-Scale Tests

Visual inspection and non-destructive examination did not reveal any cracks or crack-like indications for
any of the specimens tested as per modified NACE TM0284 (Figure 4a). Metallographic analysis - i.e.,
sectioning followed by optical and scanning electron microscopic examination - confirmed the non
destructive examination results (Figure 5). Specimens tested in NACE solution A exhibited numerous
cracks and external blistering in both parent material and weldments, as showed by SAM and
metallographic examination (Figure 4b). crack surface ration (CSR), crack length ratio (CLR), and crack
thickness ratio (CTR) values are given in Table 3 below and are well above NACE MR0175/ ISO 15156
acceptance criteria (i.e., CLR ≤ 15%, CTR ≤ 5% and CSR ≤ 2%).

a) Metallographic Examination of a b) Metallographic Examination of a

Specimen Tested in 5 wt% NaCl Specimen Tested in NACE Solution A
Solution With 10 mbar H2S for a (Weldment)
Duration of 3 Months
Figure 4: Visual and Metallographic Evaluation of specimens subjected to NACE TM0284

90 degrees (Specimen 1) 90 degrees (Specimen 1)

90 degrees (Specimen 2) 90 degrees (Specimen 1)

Weldment Weldment

a) Specimens Exposed to 10 mbar H2S for a b) Specimen Exposed to NACE Solution A

Duration of 3 months
Figure 5: SAM scans
 Table 3
Evaluation of Small Scale Specimens as per NACE TM0284

Sample ID % CSR % CLR % CTR

Parent Material 90° - Location 1 - Specimen 1 1.71 86.86 14.05
Parent Material 90° - Location 1 - Specimen 2 3.70 88.51 16.76
Parent Material 90° - Location 1 - Specimen 3 1.77 57.15 11.11
Parent Material 90° - Location 2 - Specimen 1 1.77 66.43 12.16
Parent Material 90° - Location 2 - Specimen 2 0.79 48.70 5.72
Parent Material 90° - Location 2 - Specimen 3 1.99 44.60 13.86
Weldment - Specimen 1 1.54 94.04 19.70
Weldment - Specimen 2 1.77 37.39 15.69
Weldment - Specimen 3 1.31 58.91 25.42

Large Scale Testing

Hydrogen flux reached equilibrium after 14 days exposure.

No indication of damage was found by UT examination of the test pipe prior to testing. The results of
large scale testing are summarised in Table 4 and described below.

The first crack indications were detected by UT after several weeks on the pipe section exposed to 10
mbar H2S. Whilst the number of indications increased steadily through time, less than 10% of the
indications observed showed an increase of crack area with time. This may however be an artifact
related to the imprecise nature of a manual scan for which superposition of ‘old’ and ‘new’ cracks
cannot be excluded. Depth of all indications was located around the centre line, i.e. at about 10 mm
through thickness. On completion of the test, a total of 60 crack indications with lengths ranging from 5
mm to 60 mm were found. No indications were found at or near the weldment, whilst clusters of
indications were mainly located between 250 mm and 1250 mm away from the weldment (Figure 6).

Figure 6: Crack Indication Map of the Pipe Section Exposed to 10 mbar H2S
(X: circumference of the pipe and Y: length of the pipe - The weldment is at 0mm)

Two pipe sections extracted from the ring test exposed to 10mbar H2S were selected for additional
ultrasonic inspection and determination of CAR. CAR results were found to be between 2.5% (Test
piece 1B) and 9.1% (Test piece 1A). No crack indications were found for the pipe exposed to 5mbar of
H2S.

AE registered some activity of significant importance on the pipe section exposed to 10 mbar of H2S
after two weeks of exposure. This activity was continuously registered throughout the test period, with
slight variation from week to week. The relevance of an event is determined by its energy and
amplitude. The energy and duration versus amplitude graph was used to establish the quality of the
data.

The use of multiple probes provides valuable information on the location of the event of significant
importance (Figure 7).

Figure 7: Example of Planar Location of the AE Events Recorded on the Pipe Section Exposed
to 10 mbar H2S (X: circumference; Y: length with weldment at 0mm)

A good agreement was found between AE and UT, as shown in Table 4. AE monitoring proved to be
valuable for helping target subsequent UT scans, since general crack formation was detected reliably
and even crack location was relatively precisely indicated.

As would be expected, AE provided indication of activity with no quantification of damage. In contrast,

UT provided quantitative evidence of damage. It is apparent that the two techniques are
complementary.
 Table 4
Non Destructive Examination Results

Exposure Inspection
10 mbar (0.15 psi) H2S 5.0 mbar (0.07 psi) H2S
time Method
UT - -
1 month Significant activity starting after
AE No significant activity
2 weeks exposure
UT 13 indications No indications
2 months
AE Significant activity No significant activity
UT 18 indications No indications
3 months
AE Significant activity No significant activity
UT 27 indications No indications
4 months
AE Significant activity No significant activity
UT 33 indications No indications
5 months
AE Significant activity No significant activity
59 indications after access
UT No indications
6 months constraints lifted
AE Significant increase in activity Small activity Recorded

Sectioning.

Test coupons 1A and 1B were sectioned as per NACE TM0284 and the presence of cracks was
confirmed by metallographic examination (Figure 8). The CSR, CLR and CTR are given in Table 5.
Although NACE MR0175/ISO 15156 acceptance criteria are given for testing in Solution A, it is noted
that the ‘heritage’ steel exposed to a less aggressive solution (i.e., 10 mbar H2S) fail to meet the
maximum acceptable values (Note that NACE MR0175/ISO 15156 Part 2 presently indicates a ‘no
cracking’ acceptance criteria for less severe environments that the standard ‘Solution A’).

a) Non-etched b) Etched
Figure 8: Micrograph Showing HIC Damage in Pipe Ring Exposed to 10 mbar of H2S (Test
coupon 1B)
 Table 5
Evaluation according to NACE TM0284
Test
Coupon CLR, % CTR, % CSR, %
conditions
1A-1 94.80 14.21 7.58
1A-2 99.70 5.68 5.67
1A-3a 20.29 6.00 0.56
10 mbar H2S 1A-3b 54.20 2.37 1.28
1A-4 69.90 2.68 1.88
1B-1a 48.15 8.31 4.00
1B-1b 27.26 5.79 0.61

It should be noted that the evaluation results in Table 5 was carried out on the two coupons extracted
from the areas of severe HIC, the found crack evaluating figures clearly overrate the real average
values for the tested material.
A preliminary HIC domain diagram for ‘Heritage Materials’ is plotted in Figure 9 together with data
published by the ERPG. The ‘acceptable’ region (i.e., region in which the risk of HIC is minimal) is in
dashed-green and the ‘non-acceptable’ region (i.e., region in which HIC is highly likely to occur) is in
dotted-red. The HIC behavior in the region in blank in ‘unknown’ (i.e., not tested).

6.5 6.5

6 6

5.5 5.5

5 5
pH

pH

4.5 4.5

4 4

3.5 3.5

3 3

2.5 2.5
0.001 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000
H2S partial pressure, kPa H2S partial pressure, kPa

a) ‘Heritage’ steels tested in the present study b) Modern steels

Figure 9: Preliminary HIC domain diagram for Heritage Materials

DISCUSSION

Safe Operation Threshold for Heritage Steels

The present study has indicated that ‘heritage sweet’ steels are generally more susceptible to HIC than
more recently fabricated steels. Conventional steels manufactured decades ago are significantly less
‘cleaner’ and are thus potentially more susceptible than recently fabricated [Link] hydrogen
results from the interaction between the environment and the steel, which can diffuse into a material
and collect at inclusions or other discontinuities, which are traps for H0 and where the nascent hydrogen
can re-combine to form stable molecular hydrogen. The H2 molecule is too large to diffuse through the
metal and thus H2 is trapped. The H2 gas can then accumulate, resulting in a pressure being exerted
within the metal at locations where the hydrogen gas accumulates in the traps. This may lead to off-
centred cracking at non-metallic inclusions or laminar cracking in the plane at mid-section. If such
cracks occurred on multiple planes, these cracks can then link-up and result in through thickness
leakage (step-wise-cracking).

Susceptibility to HIC is mainly related to the microstructure and cleanliness of the steel (number of trap
sites) and the hydrogen charging severity of the environment. These parameters determine the
magnitude of hydrogen flux and the extent of internal hydrogen at trap sites. Local variation of the
microstructure, resulting in a macroscopic ‘banding’ effect, may lead to the hardness increasing locally,
providing additional traps and enhancing the susceptibility at the micro-scale.

This work has demonstrated that ‘heritage sweet’ steels may not be suitable for exposure to H2S partial
pressures in the associated gas of 1 kPa (0.01 bar or 0.145 psi) and above such that existing ‘sweet’
facilities with normal operating conditions above this threshold may be at risk. ‘Heritage’ steels were
proven to be potentially more susceptible to HIC than the more recently fabricated steels tested such as
those by the EPRG and the preliminary threshold of 1 kPa (0.01 bar or 0.145 psi) suggested by their
work is therefore probably not suitable for older pipelines.

While continuous or cumulative exposure of the tested steel to 1 kPa (0.01 bar or 0.145 psi) of H2S
partial pressure for 1 month or more poses a credible threat, short-term exposure, i.e. less than 1 week
(continuous or cumulative), may be tolerable. This is supported by the acoustic emission results
showing no sign of cracking before the second week of testing.

The steel used in the present study was an API 5L X65 steel grade that had been fabricated decades
ago and has been in service for over 30 years. This material, containing 0.005 wt% sulphur and
exhibiting a ‘banded’ microstructure, has been demonstrated to be susceptible to HIC at 10 mbar of
H2S, whilst exhibiting good HIC resistance at 0.5 kPa (0.005 bar or 0.07 psi). Small variations of the
chemistry or microstructure could however impair the resistance to HIC and materials with a sulphur
content greater than 0.005 wt% might have significantly less HIC resistance. Further work is needed to
establish whether the observed HIC performance is specific to the tested material or applicable to all
‘heritage’ steels. Work is currently ongoing to establish a safe operation limit for ‘heritage’ steels at
various pH values and partial pressures of H2S and for steels with sulphur contents ranging from
0.008 wt% to less than 0.003 wt%.

As indicated, the environment affects the corrosion rate of the steel, and thus the HIC resistance.
Generally, the higher the corrosion rate, the more hydrogen that is formed. The corrosion rate is
dependent on many factors including water chemistry (especially pH), levels of CO 2/H2S, temperature,
flow rate, etc. In the present study, all tests were carried out at pH 5.2. A lower HIC resistance is likely
at a lower pH. Further work is ongoing to better define the HIC threshold at lower pH levels.

Validity of Test Methods For Heritage Steels

Whilst the HIC susceptibility of this material when exposed to 1 kPa (0.01 bar or 0.145 psi) H2S has
been demonstrated by testing a full ring section of pipe, the material successfully passed the modified
NACE TM0284 test under the same conditions. The discrepancy between the NACE Modified (i.e., mild
solution) and ‘full ring’ teats is not fully understood to date and poses some question the validity of the
modified NACE TM0284 testing method.

The material susceptibility to HIC has been demonstrated in accordance with NACE TM0284 in
Solution A and the severity of cracking was similar at the two locations tested, i.e. at the seam weld and
90° from the seam weld.

More severe damage was expected with the modified NACE test (Small scale) than the pipe ring test
(Large scale). The NACE TM0284 Modified is standard method, often used to demonstrate the
suitability of non-HIC resistant steels exposed to low level of H2S in service. This test is, in principle,
inherently more severe than the service conditions as specimens are fully immersed in the test solution
(six-side exposure), whilst the pipe ring test is more representative of service with only one side of the
pipe exposed to the fluid. The discrepancy is believed to be due to the sampling method.

In full scale testing, a much larger surface is exposed to the environment, thus statistically increasing
the number of potential hydrogen traps and thus increasing the possibility of initiating HIC. Full ring
testing mimicked faithfully the operation conditions and has proven to be a reliable method to establish
the fitness-for-purpose.

In the present study, the pipe exposed to 5mbar did not showed any degradation through the testing
period (i.e. 6 months) however some AE activities have been recorded in the last two weeks of testing.
This may be an early indication of micro-cracking and longer exposure time would be required to
determine the significance of this activity. The appropriate testing duration for ring testing remains to be
determined and is likely to vary depending on the partial pressure of H2S and pH. Low pH and/or partial
pressure of H2S may require to be exposed to the test environment for a period of time as long as 9
months.

Finally, little or no crack-like indications were found at, or around the weldment suggesting that the
susceptibility to HIC may vary within a plate, with the edges being less susceptible. This is believed to
be due to the forming processes.

These findings may have significant implications for the way in which HIC resistance of ‘Heritage’
materials should be evaluated in the oil and gas industry and further work is required to identify whether
a sufficiently robust sampling method is achievable for small scale testing.

CONCLUSIONS

The following conclusions can be drawn from the present study:

1. ‘Heritage’ steels are potentially more at risk of being susceptible to HIC than newly
fabricated steels.

2. A preliminary safe operation limit of 0.5 kPa (0.005 bar or 0.07psi) H2S at pH 5.2 is
suggested by the present study, but this is based on a very limited data set. Also a lower
threshold partial pressure of H2S might be anticipated at lower pH.

3. The linepipe steels tested (with a sulphur content as high as 0.005 wt%) and operating
within the region 0 of the NACE MR0175/ISO 15156 SSC diagram (under normal as
opposed to upset conditions) should not be at risk of cracking provided the pH is equal to or
higher than 5.2.

4. The NACE TM0284 sampling method for small scale testing of ‘Heritage’ steels may need to
be reviewed.

5. AE monitoring of HIC testing can reliably detect micro-cracks, long before they reach a
detectable size using conventional inspection methods (i.e., UT). AE has proven to be an
efficient early crack detection capability in laboratory condition. This work could predicate
taking AE further as a non-intrusive, global in-service inspection and monitoring technique
for sour service equipment.

ACKNOWLEDGEMENTS

The authors are grateful for the technical contribution and support of Catriona Smith (BP) and Bruce
Harvey (BP); the contribution of Tom Knox (BP), Darius Kermani (BP), John Nyholt (BP), Stanley
 Bottom (Firecroft), and Maria Lodeiro (NPL) on non-destructive examination; the contribution of Nick
Green (TTH), Tony Quinn (TTH) and Eric Bennett (NPL) on metallurgical examination; the contribution
of Phil Cooling (NPL) on testing; the contribution of Frank Dean (Hydrosteel) for hydrogen permeation
measurement and the support and contribution of Lee Smith (BP), Steven Shademan (BP) and Alan
Turnbull (NPL).

REFERENCES

1. ISO 15156/NACE MR0175, “Petroleum and natural gas industries - Materials for use in H2S-
containing environments in oil and gas production”, 2009

2. NACE TM0284, “Evaluation of pipeline and pressure vessel steels for resistance to hydrogen-
induced cracking”, 2011

3. OTI 95 635, “A test method to determine the susceptibility to cracking of linepipe steels in sour
service”, Health and Safety Executive, 1996

4. M.G. Hay and D.W. Rider, “Integrity management of HIC damaged pipeline and refinery
pressure vessel through hydrogen permeation measurements” CORROSION / 98, Paper no. 395 (San
Diego, CA: NACE International 1998)

5. T.G. Martin and G. Wilken, “Hydrogen induced cracking failure of SA Grade 70 Steel in near
neutral to high pH solution: part 1- field investigation”, CORROSION / 03, Paper no. 532 (San Diego
CA: NACE International 1998)

6. C. Majorel and X. Roumeau, “Examples of wet H2S cracking found in refining”, Journée de la
Fissuration par l’environnement en milieux pétrolier et gazier (Lyon, Fr : CEFRACOR 2011)

7. J.W. Martin, C. Bosch, J. Kittel, T. Casagne, “Hydrogen induced cracking (HIC) assessment of
low alloy steel linepipe for sour service applications – full scale testing”, Joint Technical Meeting on
Pipeline Research, Paper no. 28, (San Francisco, CA: 2011)

8 J Kittel, JW Martin, T Casagne, C Bosch, ‘Hydrogen induced cracking (HIC) - Laboratory

testing assessment of low alloy steel linepipe’, CORROSION / 08, Paper no. 110, (New Orleans, LA:
NACE International 2008)

9 ASTM E650, “Standard guide for mounting piezoelectric acoustic emission sensors“, 2007

10 ASTM E750, “Standard practice for characterizing acoustic emission instrumentation”, 2004

11 ASTM E2075, “Standard practice for verifying the consistency of AE-sensor response using
acrylic rod”, 2010

12 ASTM E1316, “Standard terminology for non-destructive examinations”, 2011

View publication stats