M. NAÏ et al.: ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
503–507
ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
ANALIZA GLAVNEGA VZROKA NAPAKE CEVI PRI
PREGREVALNIKU
Martin Naï, Jiøí Buzík, Tomá{ Létal, Pavel Lo{ák
Brno University of Technology, Faculty of Mechanical Engineering, Institute of Process Engineering, Technicka 2,
616 69 Brno, Czech Republic
nad@fme.vutbr.cz
Prejem rokopisa – received: 2016-07-15; sprejem za objavo – accepted for publication: 2016-09-15
doi:10.17222/mit.2016.204
Superheater-tube failure is listed among the major causes of a fossil-fuel-fired boiler outage. Therefore, it is necessary not only
to identify and repair it in the case of failure but also to eliminate the root cause of this problem. As there may be multiple
reasons of failure in exposed equipment such as a superheater, a thorough investigation of more than one probable cause is
usually required. This article focuses on a failure analysis of a boiler located in a chemical plant. After a leak was discovered,
several cracks on the superheater tubes were identified as its main cause. It was necessary to assess the extent of the damage,
detect the root cause and propose corrective actions. Two problematic locations with cracks were identified during the visual
inspection: the first was on the superheater-tube bends and the other was the weld joint between the superheater and the
transition pipe. As the first step, the material-microstructure and composition analyses of the tubes in these critical locations
were carried out. Even though small weaknesses were found in the microstructure, the main cause of the tube failure was not
identified. As the next probable cause, thermal-dilatation stresses were investigated using the finite-element analysis (angl.
FEA). The support system, consisting of fixed and spring supports, as well as the compensator were included in the analysis that
confirmed the thermal-dilatation stresses as the major cause of the failure. Based on the results, a new technical solution for the
supports was suggested and verified with the FEA.
Keywords: corrosion, weldment cracks, superheater supports, thermal dilatation
Napaka cevi pregrevalnika je znana kot ena najpogostej{ih napak pri kotlu na fosilna goriva. Zato jo je treba, ne le prepoznati in
v primeru okvare popraviti, pa~ pa tudi v splo{nem odpraviti vzrok za njen nastanek. V tako izpostavljenem elementu kot je
pregrevalnik, je za tovrstno napako lahko ve~ vzrokov, zato je za ugotovitev le-teh, potrebna temeljita preiskava. ^lanek je
osredoto~en na analize okvar kotla, ki se nahaja v kemi~ni tovarni. Potem, ko so odkrili pu{~anje, je bilo na pregrevalnih ceveh
ve~ razpok, ki so bile opredeljene kot glavni vzrok. Treba je bilo oceniti obseg {kode, odkriti vzrok in predlagati ukrepe za
popravilo. Med vizualnim pregledom sta bili ugotovljeni dve problemati~ni lokaciji z razpokami. Prva na zavojih cevi
pregrevalnika in druga v spoju zvara med pregrevalnikom in prehodno cevjo. Najprej je bila izvedena analiza mikrostrukture
materiala in analiza sestave cevi na kriti~nih mestih. ^eprav so bile ugotovljene pomanjkljivosti v mikrostrukturi, glavni vzrok
napake cevi ni bil ugotovljen. Naslednji mo`ni vzrok bi lahko bila termodilatacijska napetost, ki je bila raziskovana z uporabo
analize kon~nih elementov (FEA). Sistem za podporo, ki je sestavljen iz fiksnih in podpornih vzmeti pa tudi kompenzator, so
bili vklju~eni v analizo, ki je potrdila, da je toplotna dilatacijska napetost glavni vzrok napake. Na podlagi rezultatov je bila
predlagana nova tehni~na re{itev "podpor", ki jo je potrdila tudi FEA.
Klju~ne besede: korozija, razpoke pri zvarih, pregrevalnik, termodilatacija
1 INTRODUCTION
Boilers are common parts of many process units.
Since their life expectancy can be counted by decades
(usually 200,000 working hours), some minor or bigger
problems are inevitable. Regular inspections are carried
out after some time to find problematic areas and estab-
lish corrective actions to prevent failure. The boiler life
expectancy may vary based on the material used,
working conditions, boiler operation history, etc. One of
the most exposed boiler parts is the superheater, which is
one of the crucial heat-exchanger types among the heat-
exchanger applications. Its main advantage is the reduc-
tion of fuel consumption, but it is also susceptible to
various types of damage such as creep, deflection,
damage caused by environmental influences and mecha-
nical loads.
According to D. R. H. Jones1, the superheater is the
most commonly damaged part of a boiler, thus regular
inspections are necessary to check its condition. The
main damage reasons can be divided into three major
categories:
• Mechanical – caused by a higher stress and strain at a
specific location (deflection, cracks, weldment
damage);
• Corrosion – material-structure deformation due to
various corrosion mechanisms;
• Erosion – damage caused by the particles in the me-
dium flow.
H. Othman2 faced a similar problem of superheater-
tube deformation and cracks near the weldments. He
found that the main factor causing the problems was the
temperature of 520 °C, which caused temperature
dilatations. As soon as the supports did not allow
compensation of dilatation stresses, deformation and
cracks of the pipes occurred.
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Materiali in tehnologije / Materials and technology 51 (2017) 3, 503–507 503
UDK 620.1:621.7.019.1:621.184.3 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 51(3)503(2017)
2 BOILER DESCRIPTION
The analyzed boiler located in the chemical power
plant has been in operation for a long time under various
conditions, which differed from the original design
conditions. Therefore, regular inspections of critical
areas were necessary. Boiler design parameters are: a
steam production of 50 t/h, the nominal pressure of
superheated steam of 3.8 MPa and the nominal tempe-
rature of superheated steam of 375 10
15
−
+ °C. The boiler
consists of several crucial parts. The article is focused
only on superheaters (referred to as P1 and P2) and the
U-shaped transition pipe between them. During one of
the inspections, cracks and perforations were revealed in
the area of weldments near the inlet chamber of super-
heater P2 as well as cracks and leakage on the super-
heater-P2 pipe bend.
2.1 Description of superheaters
The construction of these superheaters is classic,
widely used all over the world. Both of them are similar,
having the same dimensions and being situated one
above the other. The output chamber of superheater P1 is
connected with the input chamber of superheater P2 by
the U-shaped transition pipe. Each superheater has 9
pipes in 4 rows (a total of 36 pipes), which go through
the membrane wall, with which they are connected by a
seal-weld joint (Figure 1).
The presented article is focused on superheater P2,
which was exposed to a medium with a temperature of
335 °C and a pressure of 4.0 MPa (Figure 2). Another
crucial part is in the U-shaped transition pipe where
cooling water is sprayed to achieve the required steam
properties.
2.2 Description of the supports
Superheater collectors are supported by two fixed
strap iron profiles (Figure 2).
The U-shaped transition pipe has spring supports on
three locations (Figure 3). Two of them are at the upper
part with displacements of 15 mm and 23 mm and one at
the bottom part with a displacement of 75 mm.
2.3 Damaged areas
As mentioned, there are two main problematic areas.
One of them includes the crack and the leakage on the
superheater-P2 pipe bend (Figure 4).
Another problem was the crack near the weld on the
superheater collector with the U-shaped transition pipe
(Figure 5). This problematic part had been repaired
several times in the past, which can be seen in the figure.
The structure of the material was expected to be one
of the possible root causes of the damage, mainly the
dislocations in the structure. Critical imperfections may
have been developed not only during the equipment
operation but even during its manufacture. In the follow-
ing step, two samples from the damaged areas were
taken for a metallurgical analysis to prove the dislocation
presence. The conformity with the declared chemical
composition and mechanical properties of material
1.0405 were tested. A spectrometric analysis was per-
M. NAÏ et al.: ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
504 Materiali in tehnologije / Materials and technology 51 (2017) 3, 503–507
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 3: Spring support of the transition pipe
Slika 3: Vzmet v preto~ni cevi
Figure 1: Part of superheater P1 and the membrane wall
Slika 1: Del super grelca-pregrevalnika P1 in stena membrane
Figure 2: Support of the superheater collector
Slika 2: Del zbiralnika v pregrevalniku
formed (Method – 12-MTL-5.4/07 program Fe-10) as
well as a tensile test (Method – ^SN EN ISO 6892-1B).
The tests proved that all the values were within the
limits.3
The first sample was taken from the pipe-bend area
because the visual inspection revealed a horizontal crack
(length 25 mm) on the pipe bend. This kind of crack
cannot have been caused by the bending process. How-
ever, the manufacturing process should have influenced
the material structure by lowering its corrosion resis-
tance. The purity results of the analyses proved the
material to be of a very good quality.
Afterwards, the corrosion that was found on both
surfaces of the pipe was examined. Pitting and small
sharp projections were found on the inside. Some of the
pits already had a character of small cracks (Figure 7),
caused by a combination of corrosion and stress. Based
on these findings, it was expected that the cracks had
started on the outside of the pipe, in the area where the
material was weakened by the small cracks from the
inner side. Stress and the medium flow might be import-
ant during the process of crack growth and will be
further examined.
The second tested sample was taken from the area
where the collector is welded to the U-shaped transition
pipe. The crack goes through the whole weld and is
divided into two branches in the middle of the
cross-section (Figure 8). The crack is open mainly in the
middle part and even with the microscope analyses, the
direction of the crack formation could not be determined.
There were no defects found either in the material
structure or in terms of material properties that would
directly influence the crack formation. Only some small
M. NAÏ et al.: ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
Materiali in tehnologije / Materials and technology 51 (2017) 3, 503–507 505
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 8: Weld macrosection with the crack (5× zoom)3
Slika 8: Prerez zvara z razpoko (5× pove~ava)3
Figure 6: Micrograph of the basic material (500× zoom, etched with
Nital)3
Slika 6: Posnetek osnovnega materiala (500× pove~ava, jedkano z
Nitalom)3
Figure 7: Crack located on the inner surface of the pipe (50× zoom)3
Slika 7: Razpoka, locirana na notranjostipovr{ine cevi (50×
pove~ava)3
Figure 5: Crack near weld on the superheater-P2 collector metallur-
gical analysis
Slika 5: Razpoka blizu zvara na pregrevalniku P2 metalur{ka analiza
zbiralnika
Figure 4: Leakage on superheater-P2 pipe bend
Slika 4: Pu{~anje na P2 pregrevalniku na krivini cevi
defects were found (see the black arrow in Figure 8). As
mentioned above, this weld was repaired several times in
the past so it was obvious that there was some problem
with the equipment geometry (e.g., supports locations),
stress and temperature dilatation.
To fully understand the root cause of the crack for-
mation, stress analyses were carried out.
2.4 FEM analyses
As the material analysis did not show the root cause
of the problem, finite-element analyses were used. A
shell model of the two collectors (one from each super-
heater) connected by the U-shaped transition pipe was
prepared in SolidWorks. For the analyses, we used com-
putational program ANSYS® Academic Research,
Release 14.5.
2.5 Mathematical model preparation
Steady-state thermal analyses and their combination
with static structural analyses were performed. At first,
the model geometry was imported from SolidWorks to
ANSYS Workbench and material properties were set up
for material 1.0405. In the next step, a computation
mesh, which had approximately 150,000 elements, was
created.
2.6 Simulation of current conditions
After preparing a sufficient mesh, appropriate boun-
dary conditions (BCs) were set up. These included the
standard earth gravity (g = 9,8066 m/s2), fixed supports
(at the place where the collector pipes were connected to
the membrane wall and also at the places of collector
supports), the internal pressure, the temperature and
displacements (at the place where three spring supports
were located). The last three BCs varied based on the
analysis type. Two load cases were considered:
Design conditions:
First, an analysis based on design parameters was
performed. The internal pressure (4.01 MPa) was applied
to all the surfaces as well as the temperature (334 °C).
Spring supports were replaced with a direction displace-
ment, which prevented the movement in the vertical
direction. Although the stress value was slightly higher
in the area of pipe bends, it was not high enough to
initiate a crack formation.
Operational conditions:
To examine the problematic areas, it was necessary to
apply operational conditions. Therefore, to all the sur-
faces, the internal pressure (p = 3,61 MPa) and two
different temperatures were applied because, as men-
tioned above, there is a part of the U-shaped transition
pipe where the cooling water is sprayed. This was the
part where the temperature varied. The first temperature
(Tmax = 332 °C) was used for one collector with pipes and
the second temperature (Tmin = 278 °C) was used for the
second collector and the part of the U-shaped transition
pipe. At first, all the other BCs remained unchanged but
the stresses still were not high enough. This is why one
of the BCs was changed and the spring supports were
simulated with the exact values of the direction displace-
ment (as mentioned before, these included 15 mm,
23 mm and 75 mm). The results of this analysis were
identical to our problem and they confirmed our expec-
tations that the problematic parts of this construction are
the spring supports.
As can be seen on Figure 9, the stress intensity is
crucial in the pipe bends, the exact location where the
crack and leakage were discovered.
An improper use of the supports was confirmed by a
higher stress intensity at the second problematic location
including the collector and the U-shaped transition-pipe
weld joint (Figure 10).
Based on these results, it was obvious that the prob-
lematic parts were the supports. They allow higher
values of the total displacement of the U-shaped transi-
tion pipe that induce higher moments and stresses. Also,
there is one support at the bottom part, witch behaves
like the center of rotation; thus, an additional stress is in-
duced. The other crucial elements were the fixed
supports of the collectors, located too close to the
weld-joint location.
M. NAÏ et al.: ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
506 Materiali in tehnologije / Materials and technology 51 (2017) 3, 503–507
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 9: Stress intensity for bent pipes
Slika 9: Intenzivnost napetosti na zvite/zavoje cevi
Figure 10: Stress intensity for the whole computational model
Slika 10: Intenzivnost napetosti za celoten model
2.7 Support modification
In order to avoid financial losses caused by repeated
shutdowns and repairs, multiple corrective measures
were proposed. The modification of the supports, which
led to a stress reduction at the crucial areas, was per-
formed. It was achieved by increasing the spring stiff-
ness, which influenced the BC displacement. Due to this
modification, each displacement was reduced by almost
a half of the previous value (10 mm, 20 mm and 35 mm).
Unfortunately, the results proved that the stress intensity
decreases insufficiently and the spring modification is
not sufficient to fix the problem. Another improvement
could take the form of minimizing the displacements or
removing the bottom support, identified as problematic.
This would lead to a further decrease in the stress levels
in the critical areas. However, these analyses are not
included in the article.
3 CONCLUSIONS
The paper focused on superheater damage. Cracks
and leakage were found in several crucial areas; thus,
material and FEM analyses were carried out. The
analyses proved that the main cause of the problems was
an inappropriate support of the U-shaped transition pipe.
The problem was caused by a vertical displacement and
also by the spring-support location, combined with the
fixed supports of the collectors near the weld location.
Due to the surface stresses induced in the problematic
areas, this place is the most ideal for a crack formation
based on shape discontinuities, heat treatment and
material discontinuities. Based on the analysis results, a
reduction of the support displacements and a change in
their location form the easiest and fastest way to prevent
increased values of the stresses and further crack
formation and propagation.
Acknowledgement
The authors gratefully acknowledge the financial
support provided by the Technology Agency of the
Czech Republic within the research project No.
TA13401000 "Waste-to-Energy (WtE) Competence
Centre".
4 REFERENCES
1 D. R. H. Jones, Creep failures of overheated boiler, superheater and
reformer tubes, Engineering Failure Analysis, 11 (2004), 873–893,
doi:10.1016/j.engfailanal.2004.03.001
2 H. Othman, J. Purbolaksono, B. Ahmad, Failure investigation on
deformed superheater tubes, Engineering Failure Analysis, 16
(2009), 329–339, doi:10.1016/j.engfailanal.2008.05.023
3 J. [temberk, M. Hala{, Materiálový rozbor vzorkù pøehøíváku kotle
K2, NDT servis s.r.o., 2014
M. NAÏ et al.: ROOT-CAUSE ANALYSIS OF SUPERHEATER-TUBE FAILURE
Materiali in tehnologije / Materials and technology 51 (2017) 3, 503–507 507
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS