Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Contents lists available at ScienceDirect

Case Studies in Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/csefa

Full Paper

Failure analysis of type 304 stainless steel amine exchanger sheets

T
in a gas sweetening plant
⁎
H. Panahia, A. Eslamia, , M.A. Golozara, A. Ashrafi Lalehb, M. Aryanpurb, M. Mazareib
a
Department of Materials Engineering, Isfahan University of Technology, Isfahan 84156-83111, Iran
b
Pardis Petrochemical Complex, Pars Special Economic Energy Zone, Asaloye 196-75391, Iran

AR TI CLE I NF O AB S T R A CT

Keywords: Amine exchangers are widely used in gas sweetening plants for amine solution regeneration
Amine exchanger process. This study is investigates cracking and leakage of a 304 type stainless steel plate in an
Cracking amine exchanger unit, used between absorption and stripper towers. In this regard, micro and
Gas sweetening plant macro examinations were performed on the plate, and solution was analyzed. Micro cracks were
Intergranular corrosion
observed on both sides of the plate, especially under the gasket region of the plate. Results
showed that the main reason of cracking was intergranular corrosion accompanied by stress.
High concentration of Formate in amine solution, and also high level of stress under gasket
region, had initiated the intergranular corrosion and cracking.

1. Introduction

Nowadays amine treating is one of the most common methods for removing acid gases from natural gas due to advantages like
high efficiency, high acid gas absorption rate and, recyclability [1]. Popularity of amine solutions generally comes from its recycl-
ability, therefore each equipment used for solution recycling is important from both financial and process points of view. Amine
exchanger is a vital equipment since it is placed between absorber and regenerator (stripper) tower, and it exchanges the heat
between lean and rich amine and finally prepares the solutions for absorbing and stripping process. Fig. 1 shows a simplified
schematic of the gas sweetening process, with the position of an amine exchanger unit shown in the figure [2].
Corrosion is one of the most known problems in amine gas sweetening plants, mainly due to presence of CO2 and H2S in natural
gas stream [3,4]. Carbon steels suffer general corrosion more than other types of steels, while stainless steels are more commonly
exposed to local forms of corrosion.
Leakage of MDEA amine solution occurred in a plate type amine exchanger in a gas sweeting plant. The amine exchanger
consisted of 504 pieces stainless steel plates which were fixed with a through bolt, and between each plate gaskets were used in order
to seal the plates from one another. The plates were made from ASTM 304 stainless steel, while other parts of exchanger were made
from ASTM A516 carbon steel. One side of the sheet plates were in contact with lean amine solution while the other sides were in
contact with semi-lean amine solution. Fig. 2. shows a schematic of a plate type amine exchanger unit.

2. Process description

Lean amine solution first enters the absorber tower while it is not very hot. Lean amine solution temperature increases during
absorption of acidic gases from top towards the bottom of the tower. At the bottom of the tower the lean amine solution is saturated

⁎
Corresponding author.
E-mail address: m.eslami@cc.iut.ac.ir (A. Eslami).

http://dx.doi.org/10.1016/j.csefa.2017.08.002
Received 29 June 2017; Received in revised form 21 August 2017; Accepted 23 August 2017
Available online 31 August 2017
2213-2902/ © 2017 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 1. Schematic of a amine gas sweetening unit. Red circle indicates the position of an amine exchanger unit [2].

from CO2/H2S gases (acidic gases), which is called rich amine solution. The rich amine solution is heated and then enters a stripper
tower in order to separate acidic gases from the solution, and converted to semi-lean and then to lean amine solution. When the lean
amine solution exits the stripping tower it is hot and should be cooled prior to transformation to the absorption tower. Therefore, in
this regard a heat exchanger unit is placed between the absorber and stripper tower, as shown by red circle in Fig. 1.
The working condition of the amine exchanger unit is summarized in Table 1. As can be seen from this table, lean amine solution
has higher temperature and lower pressure in comparison to semi-lean solution, which is mainly used for the stripping (regeneration)
process.

3. Experimental procedure

Specimens with different dimensions were cut from various regions of an amine exchanger sheet. The surface of samples was then
examined by visual, non-destructive Dye Penetrant Test (PT), Stereo Microscope, and Scanning Electron Microscope techniques. Also,
for metallographic examinations some samples were mounted in epoxy, grind to 1200 grit paper, polished and etched for a period of
30 sec in HCL-HNO3-H2O solution. Amine solution composition was analyzed by Ion Gas Chromatography (IGC). pH of both semi-

Fig. 2. Schematic of a plate type amine exchanger unit [5].

88
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Table 1
Amine exchanger working conditions.

Side Lean Semi-Lean

Position Inlet Outlet Inlet Outlet

Temp. °C 123 85 77 114

Pressure Bar 2.54 7.54

Fig. 3. Schematic of U-bend test specimens: a) non bolted specimen, bolted specimen [6].

Fig. 4. Digital camera image from both sides of a plate: a) brown weak scales b) black adhesive scales.

lean and lean amine solution was measured. Environmental induced cracking test was performed according to ASTM G30 standard
[6], in lean amine solution under nitrogen blanket.1 In this regard sheet samples were cut, drilled, bended and finally bolted. Two
types of U-bend specimens were used for the U-bend test: one bolted and the other non bolted. The bolted specimen in U-bend test
was used for investigating the effect of stress on crack initiation.2 Fig. 3 shows a schematic of the U-bend test specimens.

4. Failure analysis

4.1. Primary examination

Fig. 4 which shows digital camera image taken from both sides of a sheet sample of an amine exchanger unit. As can be seen from
this figure, one side of the sheet is covered with brown-light scales (Fig. 4a), which were relatively loose, and the other side is covered
with with black-dark scales (Fig. 4b), which were adhesive to the surface.
After initial inspections due to presence of surface scales on sheet samples, no cracks were observed using Stereo Microscope and
also Dye Penetrant Test (PT). Therefore the scales were removed from the sheet surface. After surface scale removal, while no cracks
were still observed by PT (Fig. 5), some were observed using the Stereo Microscope (see Fig. 5).
The micro cracks were mainly distributed at locations with maximum curvature on sheet surface (see Fig. 6). Formation of the

1
Lean amine solution is less corrosive than semi-lean amine solution [3,4], and it is a representative of less severe situation.
2
For the bolted specimen, according to ASTM G30 [6], the applied stress level at the stress level at the apex of U-bend specimen is roughly estimated to be close to
material yield stress. Considering fabrication residual stresses in addition to operational stress levels according to component spec sheet, the applied stress level for the
bolted U-bend test sounds reasonable.

89
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 5. Digital Camera image from: a) cleaned sheet; b, c) sheet exposed to penetrant and developer spray on gasket position; d, e) sheet exposed to penetrant and
developer spray on other regions.

micro cracks at these locations could be due to higher stress at these locations.
Further examination by Scanning Electron Microscope (SEM) showed that intergranular corrosion had initiated the intergranular
cracking (see Fig. 7). Qualitative analysis of corroded surface and surface cracks using SEM images showed that the intergranular
corrosion and cracking severity was almost similar on the both sides of the examined sheets, with more severity under the gasket.
Previous studies [7], have also shown that under the gasket region on heat exchanger sheets localized corrosion and cracking is more
often. This could be due to localized environment, and also higher stress at this region.

Fig. 6. Stereo Microscope image from microcracks on the surface of sheets shown by arrays near the pick of curvature.

90
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 7. SEM image from the surface of heat exchanger sheet at: a, b) under the gasket region at side of the sheet with black scales; c, d) under the gasket region at side
of the sheet with brown scales; e, f) other locations (on the wave formed surface of the plate).

Table 2
Chemical analysis of amine exchanger sheet in comparison to AISI standard (wt. %).

Sample Content

C Mn Si Cr Ni S P Fe

Exchanger sheet 0.06 0.84 0.4 18.5 8.8 0.025 0.007 Bal.
AISI 304 Standard < 0.08 ≤2 < 0.75 ≥18 ≥8 < 0.03 < 0.045 Bal.

4.2. Amine exchanger sheets

4.2.1. Chemical composition

Table 2 shows chemical composition of an amine exchanger sheet investigated in this study. As can be seen from this Table, the
chemical composition of the sheet is in accordance with AISI 304 standard.

91
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 8. Optical Microscope image from the microstructure of 304 stainless steel used as amine exchanger sheet at different magnifications.

4.2.2. Microstructure
Fig. 8 shows Optical Microscope image of microstructure of the sheet samples at different magnification. As can be seen from this
figure, the microstructure of sheet steel used in the amine exchanger consists of equiaxed austenite with some twin boundaries.
Formation of twins could have occurred during the sheet production process. These twins called deformation twins, should have been
produced during shaping process. Dislocation movement is not probable as deformation mechanism for such materials, since the
stacking fault energy (SFE) of austenite is low. Therefore twining is the predominant deformation mechanism than dislocation
movement. Although there are some reports stating that twin boundaries could encourage preferential corrosion and cracking [8],
this is not the cause of intergranular corrosion cracking in this study. Numerous slip bands can be also seen in Fig. 8. This could be an
indication of cold forming during fabrication of the heat exchanger sheets.

4.2.3. Energy Dispersive X-ray analysis

Segregation of Cr in stainless steels is one of the most detrimental events that leads to intergranular corrosion [9,10]. Therefore,
Energy Dispersive X-Ray Analysis (EDS) method was used in order to investigate occurrence of Cr depletion. As can be seen from
Fig. 9, no considerable amount of chemical composition inhomogeneity is observed at grain boundaries and also within the grains
through point and line scan analysis.

4.3. Surface scale analysis

As mentioned earlier (Section 4.1) there were two types of scales on both sides of amine exchanger sheets, i.e.: loose brown scales,
and adhesive black scales. Fig. 10 shows EDS analysis from both type of surface scales. From the analysis presented in Fig. 10, it can
be seen that scale’s almost have similar composition,3 and are mainly composed of Oxygen, Iron and Silicon. Since amine solution is
not essentially corrosive for 304 stainless [11], it could be concluded that Iron (Fe) in surface scale analysis (Fig. 10) should have
come from corrosion of carbon steel in other parts of amine exchanger made from A 516 carbon steel, and deposited on the surface of
the exchanger sheets. The reason for deposition of surface scales on exchanger sheets, could be due to the relatively low rate of
solution flow between the plates of the amine exchanger unit. Similar observations, have been reported in previous studies [11].
Other elements such as Oxygen and Silicon in surface deposit analysis (Fig. 10), could have come from the deficiency in the process
(Oxygen could enter the amine solution during shutdowns or leakage, and Silicon could come from anti-foams that are continuously
added to amine solution stream4).

3
The chromium peak in black scale analysis should be from the steel substrate, since it was adhesive and not removed from substrate prior to EDS analysis.
4
This was confirmed by field engineers.

92
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 9. a, b) Spot EDS analysis; c) Line scan EDS analysis with related graphs for Fe, Cr and Ni distribution.

93
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 10. EDS analysis from different surface scales formed on amine exchanger sheet surface: a) brown scale; b) black scale.

Fig. 11 shows X-ray diffraction analysis from surface scales formed on both sides of amine the exchanger sheets. As can be seen
from this figure, the main corrosion products formed are Fe2CO3 and Fe2(SO4)3 and FeS. The relative intensity of Fe2(SO4)3 and FeS
(sulfur base corrosion products) peaks in X-ray analysis presented in Fig. 11a (semi-lean amine solution side) is more than intensity of
Fe2CO3 peak in Fig. 11b (lean amine solution side). This could be due to higher amounts of H2S/CO2 in semi-lean amine solution than
lean amine solution. The higher sulfur containing compounds, could have caused more sulfur based corrosion products. In the
absence of H2S, Fe tends to react with CO2, resulting in carbonate base corrosion products depositing on equipment’s in contact with
lean amine solution.

4.4. Amine solution analysis

Table 3 shows pH of the amine solution taken from both sides of amine exchanger sheets. According to plant designation and
operational conditions,5 the pH values are in normal range.
Table 4 shows Ion Gas Chromatography (IGC) analysis of lean and semi-lean amine solutions. From the analysis it could be noted
that except for Formate anion that is almost three folds more than the maximum allowable limit (500 ppm), all components are in the

5
This was confirmed by plant engineers and sheet specification data.

94
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 11. XRD analysis from surface scales formed on both sides of the amine exchanger sheets: a) from surface scales in contact with semi-lean solution; b) from surface
scales in contact with lean amine solution.

Table 3
pH of amine solutions taken from amine exchanger’s
sheet .

Sample pH

Lean Amine 10.52

Semi-Lean Amine 9.23

range [4,12,13,14]. Formate anion is a corrosive agent which by producing formic acid, could corrode stainless steels [15].
Formate is one of the most known product of oxide degradation, which could be formed due to oxygen leakage in amine gas
sweetening plants [3,4,12,16]. Oxygen could enter the system by poor sealing, presence in make-up water or poor nitrogen blan-
keting. It is worthy to mention that caustic addition (for neutralization) could also be a reason for increase in Format level due to

95
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Table 4
Analysis of Semi-Lean and Lean amine solutions.

Unit Lean side Semi-Lean side Basic degradation compounds

Value

wt.% < 0.1 < 0.1 MEG

wt.% 0.1 0.1 THEED
wt.% < 0.1 < 0.1 HEOD
wt.% 0.1 0.1 Propanediol
wt.% 0.1 0.1 DEA
wt.% 0.1 0.1 HMPO
wt.% < 0.1 < 0.1 TEA
wt.% < 0.1 < 0.1 Bikinin
wt.% < 0.1 < 0.1 MMEA
wt.% < 0.1 < 0.1 DEA-F
wt.% < 0.1 < 0.1 HEI
wt.% 0.1 0.1 BHEP
wt.% < 0.1 < 0.1 HMP

HSS Anions
ppm 217 212 Acetate
ppm 1587 1468 Formate
ppm 247 238 Succinate
ppm <1 <1 Glycolate
ppm 121 106 Propianate
ppm <1 <1 Butyrate
ppm 3 3 Chloride
ppm 2 2 Nitrate
ppm <1 <1 Nitrite
ppm 2 <1 Oxalate
ppm 4 4 Sulfite
ppm 9 10 Sulfate
ppm 5 6 Phosphate
ppm 2 2 Thiocyanate
ppm 9 9 Thiosulfate

Cation content
mg/Kg <1 <1 Cr
mg/Kg 3 4 K
mg/Kg 94 79 Na
mg/Kg 4 13 Ni
mg/Kg 3 6 Fe
mg/Kg <1 <1 Mg
mg/Kg <1 <1 Ca

higher kinetics of Formate formation by addition of chemicals like NaOH [17].

4.5. Environmental cracking tests

Fig. 12 shows Optical Microscope images from the surface of specimens after the one month U-bend immersion test in lean amine
solution (as a representative of amine solutions in contact with amine exchanger plates), for bolted and non-bolted specimens. As can
be seen from this figure, the non bolted specimens just shows intergranular corrosion, while intergranular corrosion and cracking is
observed for the bolted specimens. Since the bolted specimens had higher stresses than the non-bolted specimens, this confirms the
intergranular stress assisted corrosion cracking.

5. Overall discussion

The composition of 304 stainless steel was in its standard range. However the amine solution showed higher amounts of Formate
than the standard range. Pure amine solutions are not corrosive in nature by themselves. In fact, corrosion of amine in refining
systems is mainly due to presence of CO2 and H2S in the flue gas. Presence of acidic gases, high temperature of process and also
leakage of oxygen in the system could cause pure amine degradation according to the following reaction [4]:
CO
⎯ MDEAH+HCOO−
CH3 N[CH2 CH2 OH]2 ⎯→
(MDEA) (Formatesalt) (1)
Degradation of amine is a catastrophic event in amine gas sweetening plants that could cause amine solution corrosiveness.
Grain boundary corrosion should have occurred in presence of degraded amine solution. Significant inhomogeneity of chemical
composition in 304 stainless microstructure was not confirmed by SEM analysis, however it could be noted that grain boundaries are
locations with higher activation energies. This can make them more susceptible to corrosion in corrosive media (i.e. the 304 stainless

96
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

Fig. 12. Optical Microscope image from .surface of 304 stainless steel after 1 month U-bend test in lean amine solution: a, b) non-bolted; c, d) bolted.

steel grain boundaries exposed to degraded amine solution).

Cracks had initiated from the corroded grain boundaries on amine exchanger sheets. The stress acting on amine exchanger sheets
could have facilitated the cracking process. Under the gasket region of the amine exchanger unit, higher stress and concentrated
solution could have formed. Therefore, with the mentioned stress assisted intergranular corrosion mechanism it is reasonable to see
higher density of internal corrosion and cracks forming under the gasket region.
The surface deposits on amine exchanger sheets, were not formed due to corrosion of the 304 stainless sheets in amine solution
environment. They should have been formed as a result of corrosion in other parts of the process, then possibly transported by the
flow of stream to amine exchanger sheets, and deposited on the amine exchanger sheets due to reduction in flow of amine solution
between the sheets.
In the event of future failure prevention, the following is recommended:

• Reduction of Formate level by limitation of oxygen leakage (consequently oxidative degradation) and/or by caustic addition.
• Using more resistive stainless steel grades, such as the 316L stainless steel grade.
• Controlling external sources of stress, such as the stresses rising from tightening of bolts in the amine exchanger units.
6. Conclusions

• The 304 stainless steel was not resistive to intergranular corrosion and cracking, in investigated amine solutions.
• There was no significant difference in severity of intergranular corrosion and cracking in lean amine and the semi lean amine
solutions.
• Formate anion was degraded to formic acid, and was identified as the main cause of corrosion of 304 stainless steel in the amine
exchanger unit.
• Corrosion and cracking was more severe under the gasket region of the amine exchanger. This is due to higher possibility of
localized corrosion, and also higher level of stress under the gasket region.

Acknowledgements

The authors would like to thank Isfahan University of Technology, and Pardis Petrochemical complex for their support.

References

[1] Gary JH, Handwerk GE. Petroleum Refining Technology and Economics. New York: Marcel Dekker; 2001.
[2] Nouri M, Mesbah J, Shaibani S. A verified computer model for corrosion rate prediction in a Di-Ethanol amine plant using field data. New Orleans: NACE; 2008.
p. 8417.
[3] Nielsen RB, Lewis KR, McCullough John G, Hansen DA. Controlling corrosion in amine treating plants. Houston: NACE; 1995. p. 571.
[4] Rooney PC, Dupart MS. Corrosion in alkanolamine plants: Causes and Minimization. Orlando: NACE; 2000. p. 494.
[5] http://www.systemcorp.co.th.

97
H. Panahi et al. Case Studies in Engineering Failure Analysis 9 (2017) 87–98

[6] ASTM G30. Standard Practice for Making and Using U-Bend Stress-Corrosion Test Specimens. West Conshohocken: ASTM; 1997.
[7] Hay MG, Baron JJ, Moffat TA. Elastomer-induced crevice corrosion and stress corrosion cracking of stainless steel heat exchanger plate in sour amine service.
Denver: NACE; 1996. p. 389.
[8] Hosford WF. Mechanical behavior of materials. New York: Cambridge University Press; 2010.
[9] Fontana MG. Corrosion Engineering. Singapore: McGraw-Hill; 1987.
[10] McGuire M. Stainless Steels For Design Engineers. Ohio: ASM International; 2008.
[11] Erfani A, Boroojerdi S, Dehghani A, Yarandi M. Investigation of carbon steel and stainless steel corrosion in a MEA based CO2 removal plant. Pet Coal
2015;57:48–55.
[12] Namazi F, Almasi H. Amine corrosion in gas sweetening plant: causes and minimization on real case study. Vancouver: NACE; 2016. p. 7178.
[13] Xie J, Simanzhenkov V, Santos B, Ikeda K, Davies L. Corrosion of UNS S30403 stainless steel trays in an amine unit. San Antonio: NACE; 2010. p. 10187.
[14] Mesgarian R. Corrosion management in gas treating plants (GTP’s): comparison between corrosion rate of DEA and MDEA a case study in sour gas refinery. IEOM
conf. 2014.
[15] Badea GE, Ionita D, Cret P. Corrosion and passivation of the 304 stainless steel in formic acid solutions. Mater Corros 2014;65:1103–10.
[16] Gouedard C, Picq D, Launay F, Carrette P. Amine degradation in CO2 capture. I. A review. IJGGC 2012;10:144–270.
[17] Litschewski Mj. More experiences with corrosion and fouling in a refinery amine system. Denver: NACE; 1996. p. 391.

98