Stress Corrosion Cracking
Stress Corrosion Cracking
17
Graham R. Lobley 1
1:Metallurgical Engineering Specialist, Consulting Services Department, Saudi Arabian Oil Company
E-mail: lobleygr@aramco.com.sa
ABSTRACT
Two examples of stress corrosion cracking (SCC) in refinery equipment are presented to highlight the
lessons learned to prevent recurrence of such failures. In each failure analysis case study, the
cracking is characterized using various metallurgical laboratory techniques, including metallography
and fractography, in order to identify the mode of failure. Examples of polythionic acid SCC (PASCC)
in a sensitized austenitic stainless steel and caustic SCC in carbon steel equipment are presented.
PASCC of 304 stainless steel charge heater tubes occurred in a naphtha hydrotreater during
shutdown, after about 18 years in service. Caustic SCC of carbon steel piping in an LPG Merox gas
sweetening plant occurred in non-stress-relieved downstream piping, due to caustic carryover from
the Merox unit. Cracking characteristics are discussed and options to avoid recurrence are presented.
Keywords: Failure analysis, polythionic acid, caustic, metallography, fractography.
1. INTRODUCTION
2.1. Introduction
The 304 stainless steel charge heater coil of a naphtha hydrotreater unit (NHT) was found to
be extensively cracked following unit shutdown, when soda ash washing was applied. The
NHT is fired on fuel oil/gas. The coil carries sour naphtha (350 ppmw S) feed and hydrogen
is added to the feed at a process temperature of 330 °C (see simplified process flow diagram,
Fig. 1). The unit had been in service for about 18 years and had been successfully shut down
and restarted several times previously in its lifetime.
2.2. Examination
Tube sections from the heater coil were received for laboratory evaluation. These sections
showed several transverse cracks, some with significant crack opening (Fig. 2). Beneath a
rust flash, the internal coil surfaces were coated with a fairly thick, powdery black deposit,
which was sampled for chemical analysis. Microsections taken across the cracks confirmed
that the cracking had started from the internal, process side of the tube (Fig. 3).
Metallography showed that cracking was intergranular with some branching (Figs. 4a, 4b).
An intergranular failure was also confirmed by fractography of an opened crack (Fig. 4c).
Electrolytic etching of a microsection in oxalic acid (ASTM A262 Practice A) revealed wide,
ditched grain boundaries, indicative of sensitization (Fig. 4d). X-ray diffraction (XRD)
revealed that the black deposit inside the tubes was 80% iron sulfide (FeS), balance iron oxide
hydroxide FeOOH. The tube material met the compositional requirements of the specified
ASTM A271 TP304. The alloy analysis was 18.5%Cr, 9.95%Ni, 1.55%Mn, 0.64%Si,
0.07%C.
Top
Stress Corrosion Cracking: Case Studies in Refinery Equipment Vol. 5. 19
2.3. Discussion
In the early 1960s [Samans, 1964], stressed and sensitized 304 stainless steels were first
reported to suffer intergranular stress corrosion cracking in a solution of polythionic acids
(H2SXO6), which may form in petroleum refinery units during shutdown. Only reaction of
sulfide, moisture and oxygen is necessary. In the presence of sulfide corrosion products, such
conditions are met during shutdown, when moisture originates from general humidity,
washing or steam out and oxygen as air enters the unit upon opening to atmosphere.
Subsequently, this particular form of intergranular SCC was called polythionic acid stress
corrosion cracking (PASCC). However, Piehl considered that the corrosive species actually
responsible for the intergranular SCC was probably sulfurous rather than polythionic acids
[Piehl 1964], which also forms readily during refinery shutdown conditions. Both authors
agreed that service temperatures below 800 °F (427 °C) would not result in the required
sensitization of austenitic stainless steels. More recent work [Moller 1984] has suggested that
of the polythionic acids, only tetrathionic acid (H2S406) causes SCC in sensitized 304 stainless
steel. Cracking has also been reported in refinery equipment with lower operating
temperatures. Moller [Moller 1984] cites a case of PASCC of an AISI 304 flange from a
hydrodesulfurizer. The plant had operated for 138,000 hours (15.75 years) and the flange
experienced temperatures up to 384 °C (720 °F). The 304 flange eventually became
sensitized after prolonged exposure at relatively low temperatures.
Unless suitable precautions are taken, in accordance with NACE RP0170 [NACE, 1997],
austenitic stainless steels which have been exposed to sulfur-bearing refinery process streams
within the sensitization range may become susceptible to intergranular stress corrosion
cracking. NACE RP0170 practice recommends soda ash neutralization washing and nitrogen
purging procedures (note this NACE procedure is currently under revision). Investigation
revealed that the soda ash neutralization procedures were not properly documented and
followed and that the protective alkali film was wrongly water washed at the end. An
important question is what is the lowest service temperature below which sensitization does
not occur. In our case, the process stream temperature is 330 °C, which initially appeared too
low to cause sensitization. However, the internal metal skin temperature is actually higher
(approximately 405 °C) due to conduction, since this is a fired heater.
Top
Vol. 5. 20 Graham R. Lobley
1. PASCC failure occurred in a sensitized 304 stainless steel, after 18 years’ service.
2. NACE RP0170 shutdown procedures should be carefully followed for austenitic
stainless steel equipment.
3. Upgrade the coil to AISI 347 stainless steel, which is significantly more resistant to
sensitization. This alloy should be given a stabilization heat treatment.
3.1. Introduction
In an LPG Merox system, an ASTM A105 8-inch carbon steel elbow of developed a small
leak. Merox is a gas sweetening process, which removes mercaptans by an oxidation process
and is performed in a caustic solution.
The carbon steel piping (ASTM A53 grade B) and equipment of the system (heat exchangers
and splitter) were not originally stress relieved. The failure location was apparently from the
weld, close to the 6 o’clock position, near a drain point. While preparing the original elbow
for re-welding, plant inspectors noticed that there were transverse cracks within the heat
affected zone of the weld. It was therefore decided to replace both elbows and locally stress
relieve. As the carbon steel line was not stress relieved and an incident of caustic carryover
had occurred a short time before the failure, inspection personnel began to suspect that caustic
stress cracking might be the cause of the elbow failure. The equipment had been in service
since 1972, but it was suspected that occasional caustic carryover incidents may have started
following upgrade of another refinery unit in 1995. The stream temperature downstream of
the C3/C4 splitter (Fig. 5) is 93 °C and approximately 15 wt% caustic is used in the Merox
unit.
3.2. Examination
Fig. 6 shows the portion of the elbow received for investigation. In the weld preparation area,
networks of transverse cracks were clearly visible under low power examination (Fig. 6,
bottom). Microsections were taken transversely across the cracks. Branched intergranular
cracks were noted (Fig. 7), which were filled with dark grey corrosion products. The elbow
was a low carbon steel composition (0.11%C, 0.47%Mn, 0.22% Si, 0.034%P, 0.021%S).
Energy Dispersive Spectroscopy microanalysis of the opened surfaces of one crack revealed
mainly iron and oxygen, with significant levels of sodium also present. This confirmed that
the cracks were filled mainly with iron oxides and the detection of sodium is consistent with
caustic carryover into the piping downstream of the Merox unit. These findings confirm that
caustic stress corrosion cracking was responsible for the elbow failure.
Top
Stress Corrosion Cracking: Case Studies in Refinery Equipment Vol. 5. 21
3.3. Discussion
Caustic stress corrosion cracking of steels is one of the oldest known forms of SCC, dating
back to the days of the early steam locomotives, when it was responsible for countless
explosions of riveted boilers [Trethewey et al., 1985]. Incidents of caustic cracking of low
alloy steels in nuclear and fossil-fueled power plants from 1969 onwards focused interest in
this form of SCC within the power industry [Lyle, 1983]. Within the process industries,
caustic SCC has caused numerous failures in welded steel piping and pressure vessels exposed
to caustic solutions at high temperatures. During the early 1990s, several incidents of caustic
stress corrosion cracking of carbon steel were encountered within company gas plants and
refineries [McIntyre 1994]. The NACE caustic soda service graph [NACE, 1985] indicates a
“safe” region where stress relief of carbon steels is not specified below certain operating
temperatures and caustic concentrations, but in some cases cracking was encountered at
nominal process temperatures within the safe region [McIntyre, 1994].
Several specialists have reviewed possible explanations for the intergranular crack path
typical of caustic cracking in ferritic steels. Microsegregation of certain elements may be
responsible for creating grain boundary heterogeneity, and in this case intergranular corrosion
is expected in the absence of any stress [Parkins, 1994]. This is the case for caustic and mild
steels and it has been suggested that phosphorus is the segregating element [Parkins, 1994,
Jones et al., 1995]. In SCC in high pH caustic solutions, iron dissolves as oxyanions
(HFeO2_) and magnetite (Fe3O4) is precipitated within the crack [Scully, 1966] This can be
understood by referring to the potential-pH stability diagram (Pourbaix diagram) for iron.
Electrochemistry plays an important role in cracking, as caustic SCC has been found to occur
over a fairly narrow range of electrode potential corresponding to a transition zone between a
passive magnetite film and the oxyanion [Parkins, 1994, Jones et al. 1992]. If the protective
film is ruptured, localized dissolution and SCC may occur.
Following the recent field experience within the company [McIntyre, 1994], engineering
standards have been tightened to require postweld heat treatment to stress relieve piping and
equipment in all caustic service, including conditions where caustic carryover may occur. All
new carbon and low alloy steel equipment is now therefore stress-relieved for this service, but
automatic retrospective action for all existing piping and equipment was not taken, except in
selected cases. In considering inspection for SCC, wet fluorescent magnetic particle testing
(WFMPT) was recommended as the most sensitive and preferred method [McIntyre, 1994].
WFMPT requires access to the process side surfaces. As cracking is invariably near welds,
external inspection methods such as ultrasonic or radiography testing are more difficult to
interpret, since defect indications can include real SCC cracks as well as lack of weld fusion.
Top
Vol. 5. 22 Graham R. Lobley
1. Caustic stress corrosion cracking was confirmed within the heat affected zone of the
carbon steel elbow.
2. Internally inspect both the splitter and reboiler by WFMPT at the next shutdown.
3. Plan to replace the present piping with stress-relieved piping during the next
shutdown.
4. SUMMARY
SCC of the types discussed can be prevented by optimal material specification and by
carefully following established mitigation procedures. Contractor’s on-site work, e.g. soda
ash washing treatments, should also be actively monitored to ensure compliance with
approved practices.
ACKNOWLEDGEMENTS
REFERENCES
1. Hamner, N. E., 1985, Corrosion Data Survey, Metals Section, 6th Edition, NACE, p.176.
2. Jones, R. H., Ricker, R.E., 1992 “Mechanisms of SCC” Chapter in Stress Corrosion Cracking,
ASM
3. Lyle, F. F., 1983, “Corrosion Cracking Characterization of 3.5 NiCrMoV Turbine Rotor Steels in
NaOH and NaCl Solutions”, Corrosion, vol. 39, 120 - 131, 1983.
4. McIntyre D. R., 1994, “ Review of Caustic Stress Corrosion Cracking and Its Occurrence in Saudi
Aramco Plants” Proceedings of the Materials Engineering and Corrosion Control Technical
Exchange Meeting, Saudi Arabian Oil Company, Dhahran, December 3 – 5.
5. Moller, G. E, 1984, “Designing with Stainless Steels in Stress Corrosion Environments”,
Corrosion Source Book, ASM, 266 - 277.
6. NACE RP0170-97, 1997: “Protection of Austenitic Stainless Steels and Other Austenitic Alloys
from Polythionic Acid Stress Corrosion Cracking During Shutdown of Refinery Equipment”.
7. Parkins, R. N., Stress-corrosion Cracking of Ferritic Steels, 8:33 – 8:50, in Corrosion Vol. 1
Metal Environment Reactions, Butterworth Heineman, (Eds. L. L Shrier, R. A. Jarman,
G. T. Burstein), 1994.
8. Piehl, R. L., “Stress Corrosion Cracking by Sulfur Acids”, Proc. API, vol. 44[III], 189 - 197,
1964.
Top
Stress Corrosion Cracking: Case Studies in Refinery Equipment Vol. 5. 23
9. Samans, C. H., 1964, “Stress Corrosion Cracking Susceptibility of Stainless Steels and Nickel-
base Alloys In Polythionic Acid and Acid Copper Sulfate Solution”, Corrosion, vol. 20,
236t – 262t.
10. Scully, J. C., 1966, The Fundamentals of Corrosion, Pergamon, p. 163 –164.
11. Sedriks, A. J., 1992, “SCC of Stainless Steels” Chapter in Stress Corrosion Cracking, ASM.
12. Trethewey, K. R., Chamberlain, J., 1995, Corrosion for Science and Engineering, Longman.
Fired Reactor
Heater
Untreated
Naphtha (330 °C)
S 350 ppmw
Fuel Gas/
H2 Fuel Oil Naphtha
(S <0.5 ppmw)
+ H2S
Top
Vol. 5. 24 Graham R. Lobley
x16 x100
x750 x320
Fig. 4
(a) intergranular cracking (b) intergranular cracking
(c) intergranular cracking (d) sensitized material
Top
Stress Corrosion Cracking: Case Studies in Refinery Equipment Vol. 5. 25
Butane (C4)
to storage Propane (C3)
40 ºC
LPG from
Feed/Bottoms C3/C4 Splitter
Merox Exchanger
Splitter
Reboiler
93 ºC
Top
Vol. 5. 26 Graham R. Lobley
Fig. 7: microsections
top unetched x16
centre etched x32
bottom x400
Top