PAGE1OF31
PT PUPUK SRIWIDJAJA (PERSERO)REALIBILITY AND QUALITY ASSURANCE DEPARTMENT FIELD TECHNICAL INSPECTION – II
MEMO TO FILE
NUMBER : 001/MTF/ITL-PII/2010
AUXILIARY BOILER2A – 101BUPUSRI - II
BY:
MAULIDIN BASTIAN
On behalf of Failure Analysis team of:MAULIDIN BASTIAN/MARTHA INDRIYATI /DIKDIK YULIANA / BHARATA
andFIELD TECHNICAL INSPECTION COORDINATOR-I
PAGE2OF31
PT PUPUK SRIWIDJAJA (PERSERO)REALIBILITY AND QUALITY ASSURANCE DEPARTMENT FIELD TECHNICAL INSPECTION – II
MEMO TO FILE
NUMBER : 001/MTF/ITL-PII/2010
AUXILIARY BOILER2A – 101BUPUSRI - II
Prepared By, Agreed By, To be known by,
Maulidin Bastian M. Toni A. Muksin
PAGE3OF31
CONTENTS
A. ABSTRACT............................................................................................................................5
B. INTRODUCTION....................................................................................................................5
C. OBJECTIVE............................................................................................................................6
D. CHRONOLOGIES..................................................................................................................6
E. LITERATURE REVIEW........................................................................................................8
1. AUXILIARY HEATER DESIGN.............................................................................................8
2. STRESS RUPTURE OR SHORT-TERM OVERHEATING.............................................11
3. METALLURGICAL AFFECTS.............................................................................................11
4. CHEMICAL ANALYSIS........................................................................................................13
5. MECHANICAL PROPERTIES.............................................................................................13
F. METHODOLOGY.................................................................................................................16
G. RESULTS AND DISCUSSION..........................................................................................18
1. START UP AND FAILURE PROCESS..............................................................................18
2. MECHANICAL PROPERTIES.............................................................................................20
3. PHYSICAL PROPERTIES...................................................................................................22
4. METALLURGICAL EFFECTS.............................................................................................25
5. THERMAL CRACKS.............................................................................................................28
H. CONCLUSION.....................................................................................................................29
I. Bibliography........................................................................................................................30
PAGE4OF31
LIST OF FIGURES
Figure 1. Tube impose a reaction force as in inside Primary Reformer Furnace at Auxiliary
Heater 2A-101BU
Figure 2.Variation of fluid temperature and tube-wall temperature as water is heated through
the boiling point with low, moderate, high, and very high heat fluxes (rates of heat
transfer)(Steam, 1972)
Figure 3.Thin-lip rupture in a boiler tube that was caused by rapid overheating. This rupture
exhibits a “cobra” appearance as a result of lateral bending under the reaction
force imposed by escaping steam. The tube was a 64-mm outside-diameter × 6.4-
mm (0.250-in.) wall thickness boiler tube made of 1.25Cr-0.5Mo steel (ASME SA-
213, grade T-11).
Figure 4. A compilation of ultimate tensile strength versus Brinell hardness number for
selected metals based on handbook data.
Figure 5. Location of failure in Auxiliary Boiler
Figure 6. Evidences that found on header which can be blockaging media for steam flow
(welding debris, rod, plate, etc)
Figure 7. Relationship between Ultimate and Yield Strength with Temperature
Figure 8. The thickness measurement in failure tube
Figure 9. Cobra lip in Auxiliary Boiler 101BU Pusri 2 tube failure
Figure 10. The crack phenomenon which shown after failure
Figure 11. (a) Microstructure of crack lip ; (b) macrostructure of crack lip
Figure 12. The place of SA 106 Gr.B in Fe-C Diagram
Figure 13. Crack happened in weld joint of bottom header at no.1 from East and no.3 from
South
Figure 14. The mechanism of thermal crack
LIST OF TABLES
Table 1. Material SA 106 Gr.B specification
Table 2. Chemical composition of material SA 106 Gr.B
Table 3. Chronologies of Auxiliary Boiler tube failure
Table 4. Tensile strength properties of SA 106 Gr.B
Table 5. Hardness Measurement for each samples
Table 6. Strength Properties of SA 106 Gr.B material
PAGE5OF31
Table 7. Conversion of Hardness vs Tensile Strength for tube after failure in each location
Table 8. Thickness Measurement for failure tube
A. ABSTRACT
In allusion to the explosion of Auxiliary Boiler took place at PUSRI II in Turn Around
Plant August 2010 having a series of evidence collection and inspection jobs which
includes collecting operation condition and parameters, bent-tube, sampling the
explosion fracture, fin-tube apart from explosion fracture and fracture detection
medium and its hangover, etc., had been carried out firstly. Based on these jobs,
farther analysis and computation work has been done to the structural and materials
characteristics and the operation condition of the auxiliary boiler, including
composition, metallographic phases, complemented by chemical analyses of tube or
fireside deposits, tensile properties, impact energy, strain ageing characteristics and
fracture toughness of the fin-tube steels, the preview of leak detection medium and
it’s hangover in the boiler and also explosion energy, as appropriate,. The most
probable cause of the unexpected tube failure is the major factor causing unreliability
in boilers. Characterizing the degree of microstructural degradation can also help to
confirm and separate various potential high temperature tube damage modes. The
main medium reason which caused short-term overheating failure mode is abnormal
condition in presence of tube blockaging. This project is also investigating failure
mode which was happened in Primary Reformer Furnace at Auxiliary Heater(or
commonly, Auxiliary Boiler) 2A-101BU, PT Pupuk Sriwidjaja, Palembang-Indonesia,
fin-tube at start-up process and sharing ideas for improving the process and turn-
around plant.
B. INTRODUCTION
When water is boiled in a tube having uniform heat flux (rate of heat transfer) along
its length under conditions that produce a state of dynamic equilibrium, various
points along the tube will be in contact with sub-cooled water, boiling water, low-
quality steam, high-quality steam, and superheated steam. A temperature gradient
between the tube wall and the fluid within the tube provides the driving force for heat
transfer at any point.
PAGE6OF31
Port & Herro assigned a brief explanation of failures caused by short-term
overheating. Short-term overheating occurs when the tube temperature rises above
design limits for a brief period a boiler operation upset. Conditions leading to short-
term overheating, generally are partial or total tube plug gage and insufficient coolant
flow due to upset conditions and/or excessive fire-side heat input.
C. OBJECTIVE
This project is investigating failure mode which was happened in Primary Reformer
Furnace at Auxiliary Boiler 2A-101BU fin-tube at start-up process and sharing ideas
for improving the process and turn-around plant.
D. CHRONOLOGIES
A water leak was detected in Primary Reformer Furnace at Auxiliary Heater 2A-
101BU section on nine months after the boiler had been maintained. The failure was
located in farthest section from south side burner, clearly at MK “E” header coil.
BENT TUBE
PAGE7OF31
Figure 1. Tube impose a reaction force as in inside Primary Reformer Furnace at Auxiliary Heater 2A-101BU
The MK-E (coil) steam generating system is finned tube which having initial
properties as follows,
OD Wall Thickness SCH Material Type
48,26mm (1,9”) 0.508 cm (0,2”) 80 SA-106 Gr B Fin-Tube
Table 1. Material SA 106 Gr.B specification
The tube section was placed into the boiler that makes steam production around
5193 BTU/hr.ft2 as shown in Figure 1. The chemical composition of tube material is,
Material %C %P %Cr %Mo %S Application
SA106 Gr.B .[4]0,3
(mid-CS)
0.035
(max)0.4 0.15
0.035
(max)Boiler, etc
Table 2 Chemical composition of material SA 106 Gr.B
The fin tube is helically wound serrated with subsequently either annealed or
normalized and 112 tubes attached together. The coil can be operated at 1500 psig
(105 kg/cm2G) in inlet and also outlet with 325°C (598K) steam pressure, providing
an equivalent output in coil of 102,000,000 BTU per Hr.
On the failure time, operating parameters happened at;
Date and Time Explanations
August 12nd, 2010 Start-up of Ammonia Plant with 85% Gas Rate.
14.00 WIB Vent gas in PIC-8
Steam drum Steam outlet = 150 tonhr
BFW flow to steam drum (FRA-26) =195 tonhr
Steam drum Level (LRA-24) = 65 %.
HS Pressure in steam drum (PRC-18) = 80 kg/cm2.
PAGE8OF31
Discharge Pressure 104-JA Pump to Steam Drum = 95
kg/cm2.
Four (4 Ea) Main burner in Auxiliary Boiler was firing at 75%
average load.
Flue Gas Temperature = 650°C
101B Riser Temperature = 760°C
20.30 WIB Increasing of Gas Supply Rate from 80 to 85%
21.30 WIB Auxiliary Boiler blown up suddenly and fire was detected from
Main Burner No 1 & 2 ( From Bottom Side)
PIC-21 data shown the changes of pressure from -0.9mmHg
(Normal Pressure) to +1.5mmHg(Positive Pressure)
I.D Fan 101-BJT Speed increased from normal condition
(4200RPM) to 5400 RPM
Fuel Gas from Battery Limit was being cut to shut down the
plant
August 13th, 2010 1 tube was failure and bending in MK-E Header Coil. Failure
location was detected at Tube no.2 from East and 4th Row
from South
1 crack tube was found in weld joint tube-top Header and
located at no.1 from East and no.4 from south
1 leak was found in weld joint of bottom header and located at
no.1 from East and no.3 from South
Tube located at no.1, 2 and 3 from East and no.4 from South (
3Ea) was being plugged
Tube located at no.2 from East and no.3 from South ( 1Ea)
was being plugged too.
August 14th, 2010 Fail Tube was delivered to NDT-Lab for investigating the
failure mode.
Table 3. Chronologies of Auxiliary Boiler tube failure
E. LITERATURE REVIEW
PAGE9OF31
1. AUXILIARY HEATER DESIGN
The design of a steam-generating unit balances the heat input from the combustion
of a fossil fuel with the formation and superheating of steam. The heat absorbed is
converted into steam at its saturation temperature, a function of the operating boiler
pressure. Within the convection passes, the flue-gas temperature is further reduced
by the superheating or reheating of steam in superheaters and re-heaters. To extract
more heat and to improve overall thermal efficiency, an economizer preheats the
boiler feedwater to a temperature close to its boiling point. The flue gas travels
through an air preheater, which heats the combustion air, then makes its way up the
stack.
The formula for steady-state heat transfer is,
QA0
=U 0∆T❑
…. Formula 1
where QA0
(in Btuhrft 2) is the heat flux per unit area, U0 (in
Btuhrft 2° F) is the overall
heat-transfer coefficient, and ΔT (in 0F) is the temperature difference that drives the
heat flow.
There are unintended conditions that drive the system to another analysis which
using the equation of individual thermal resistances for heat flow. And the system will
run from flue gas to steam and also individual temperature gradient, the net effect
will be an increase in tube metal temperature. In simplify, decreases in steam-side
heat-transfer coefficient caused by reduced flow, will lead to tube metal temperature
increases.
PAGE10OF31
Figure 2.Variation of fluid temperature and tube-wall temperature as water is heated through the boiling point with low, moderate, high, and very high heat fluxes (rates of heat transfer) (Steam, 1972)
Figure 2 indicates the effects that different heat fluxes have on tube-wall
temperature. In the region where sub-cooled water contacts the tube (at left, Fig. 2),
the resistance of the fluid film is relatively low; therefore, a small temperature
difference sustains heat transfer at all heat-flux levels. However, the resistance of a
vapor film in steam of low quality is relatively high; therefore, at the onset of film
boiling, a large temperature difference between the tube wall and the bulk fluid is
required to sustain a high heat flux across the film. The effect of the onset of film
boiling on tube-wall temperature appears as sharp breaks in the curves for
moderate, high, and very high heat fluxes in Fig. 2. With increasing heat flux, the
onset of unstable film boiling, also known as departure from nucleate boiling (DNB),
occurs at lower steam qualities, and tube-wall temperatures reach higher peak
values before stable film boiling, which requires a lower temperature difference to
sustain a given heat flux, is established.
PAGE11OF31
2. STRESS RUPTURE OR SHORT-TERM OVERHEATING
Short term overheating failure is one in which a single incident or a small number of
incidents exposes the tube steel to an excessively high temperature (hundreds of
degrees above normal) to the point where deformation or yielding occurs. The
conditions such as loss of coolant flow and excessive boiler gas temperature are
happened frequently in this failure.
Viswanathan stated abnormal conditions that driven to failure as:
Internal blockage of tube
Loss of coolant circulation or low water level
Loss of coolant due to an upstream tube failure
Overfiring or uneven firing or boiler fuel burners
The first three produce starvation and the tube can be blocked by erection or repair
debris, tools, steel shot, pre-boiler oxides, deposits from carryover or spray water, or
loose pieces of internal non-pressure part hardware such as scrap plates, bolts and
nuts which driven the worst condition like approach the furnace-gas temperature.
3. METALLURGICAL AFFECTS
In general, short-term overheating involves considerable tube deformation in the
form of metal elongation and reduction in wall area or cross section. Such failures
indicate wall thinning and local bulging precede the actual fracture, because the
strength of the material is reduced at the higher temperature.
Hence, for metallographic explanation, a fishmouth appearance with thin-edge
fracture surfaces and considerable swelling is typical for a ferritic steel tube that has
failed before its temperature has exceeded the upper critical temperature ( Ac3 ¿.
Figure if, however, the tube temperature was high enough to transform the iron in the
steel from ferrite to austenite, there will be no noticeable “necking down,” or
reduction in wall thickness, of the fracture edges. Microstructure of the steel will be
performed to confirm that the tube temperature prior to failure was high enough to
transform the ferrite to austenite.
PAGE12OF31
French is explained about specific indications on failure caused by rapid overheating.
The thin-lip, common shapes which happens in short-term ruptures are usually
transgranular tensile fractures occurring at metal temperatures from 650 to 870 °C
(1200 to 1600 °F). These elevated-temperature tensile fractures exhibit macroscopic
and microscopic features that are characteristics of the tube alloy and the
temperature at which rupture occurred. A tensile fracture results from rapid
overheating to a temperature considerably above the safe working temperature for
the tube material and is accompanied by considerable swelling of the tube in the
regions adjacent to the rupture that have been exposed to the highest temperatures.
As shown in Fig. 3, steam escaping at high velocity through the rupture will
sometimes impose a reaction force on the tube that is sufficient to bend it laterally.
Figure 3.Thin-lip rupture in a boiler tube that was caused by rapid overheating. This rupture exhibits a “cobra” appearance as a result of lateral bending under the reaction force imposed by escaping steam. The tube was a 64-mm outside-diameter × 6.4-mm (0.250-in.) wall thickness boiler tube made of 1.25Cr-0.5Mo steel (ASME SA-213, grade T-11).
PAGE13OF31
Changes in tube ID and OD measurements can be indicators of overheating.
Increases of 5% or more are indicative of short-term overheating. Also, significant
microstructural changes in carbon steel will occur when the steel is overheated, and
these changes can be used to estimate the metal temperature at failure.
4. CHEMICAL ANALYSIS
Chemical analysis should be done prior to before-after exposed to overheating and
also make carbon equivalent calculation to predict phase happening.
Formula 2
If the relative amounts of ferrite and martensite can be determined by microstructural
analysis, and if the alloy composition is known, a Fe-C equilibrium phase diagram
can be used to estimate the metal temperature at the time the tube burst.
Chemical analysis is subjected to the analysis tool for checking before-after material
condition. If there is any chemical intrusion during failure, physically it will appear as
deposit. This analysis also should be done for assuring external effect which come to
form of tissue layer or deposit on tube or fireside.
5. MECHANICAL PROPERTIES
Chemical analyses of tube or fireside deposits will continue to tensile properties
checking, impact energy, strain ageing characteristics and fracture toughness of the
fin-tube steels, the preview of leak detection medium and it’s hangover in the boiler
and also explosion energy,
An approximate relationship between the hardness and the tensile strength (of steel) is,
TS(Mpa)=3.55*HB (HB=<175) else 3.38*HB (HB>175) … Formula 3
TS(Psi)=515*HB (HB=<175) else 490*HB (HB>175) … Formula 4
whereHB is the Brinnell Hardness of the material, as measured with a standard
indenter and a 3000 kgf load.
PAGE14OF31
Table 4. Tensile strength properties of SA 106 Gr.B
PAGE15OF31
Figure 4. A compilation of ultimate tensile strength versus Brinell hardness number for selected metals based on handbook data.
A comparison of ultimate tensile strength with Brinell hardness is shown in Fig. 4 for
several metals. Many of the copper-base alloys have much higher UTS than
predicted and the gray irons tend to have a much lower UTS than that predicted. In
the case of gray iron, the graphite flakes serve as crack initiators in tension, whereas
these flakes are under compression during the hardness test, and work hardening
that occurs increases the hardness. Ductile irons also are limited in tensile strain, or
elongation, but not as severely as gray irons because the graphite is nodular rather
than flake. Desulfurized steel would behave in a manner similar to that of ductile vs.
gray irons, when compared with plain carbon steel.
To analyze tension characteristic of recently pipes in resist transversal tension
causing from leakage, the formula of Tension vs(P, Di, t) will be using as follows
σ tr=P Di2 t
… Formula 5
PAGE16OF31
F. METHODOLOGY
A 30.5 cm (1 ft) long piece of the failed tube was cut out for analysis and examined
metallographically. The failed piece was sectioned, and regions associated with
crack initiation were examined under the optical stereoscope and microscope.
Chemical analysis will be done using X-Met to reveal information for predicting
material deposit and at elevated temperature using carbon equivalent calculation for
predicting phase happening.
Mechanical properties calculations which are giving advantage to stress to failure
calculation are headed to hardness measurement at bottom side of lip, at lip,
opening crack and bare tube.
Samples tracking :
A = Inside tube (farthest location from failure)
B = Circumferential side (farthest location from failure)
C = Location of failure which is not breaking and perpendicular to alligator lip
D = Location of failure which have transition to deform plastically
E = Location of failure which deformed plastically
PAGE17OF31
Figure 4. Thin-lip rupture in a boiler tube that was caused by rapid overheating happened to fin tube Primary Reformer Furnace at Auxiliary Section (102B) PUSRI II Plant
Sample A
Sample B
Sample C
Sample D
Sample E
PAGE18OF31
PAGE19OF31
G. RESULTS AND DISCUSSION
1. START UP AND FAILURE PROCESS
In the MK “E” region of Auxiliary Boiler, 101-BU (Figure 6), the fin-tube was attached
to header and steam flow was delivered from down to upper side of auxiliary boiler.
In firing there was an accident where Gas Supply Rate increased from 80-85% and
Riser Temperature reached ±750°C. This temperature was fit enough for material
SA-106 Gr.B to be damaged and got short-term rupture. For the failure criteria itself,
the temperature between 650-800°C should make yielding phenomenon due to
mechanical properties degradation and phase changes on the fin-tube. The weakest
section will be failed after explosion to be more-yielding than creep and if the other
tube surround got rapid increasing of pressure and temperature, crack will happened
in joint between header and tube.
The failure location in the farthest section was found because of lack of steam
distribution on that side where it should be fitted with relevant steam condition in all
tubes. Figure 5 stated the location of failure as follow,
Figure 5. Location of failure in Auxiliary Boiler
By field, after the accident tube was examined
with Boroscope to reveal information inside tube
and header. The surprising evidences that found
on the results were blockaging media for steam
flow is located on the header and overfiring or
uneven firing on boiler fuel burners shown in the
figure 8 below. Viswanathan stated the
abnormal condition as major cause of failure
happened.
PAGE20OF31
(a) (b) (c)
(d) (e) (f)
Figure 6. Evidences that found on header which can be blockaging media for steam flow (welding debris, rod, plate, etc)
13 Agt 2010
Welding debris Rod Bent Rod
Plate
PAGE21OF31
Based on Figure 6 above, the unintended objects caused change of tube wall
temperature and it drive to temperature gradient and lack of heat transfer.
Consequently, decreases steam-side heat-transfer coeficient caused by reduced
flow will lead to tube metal temperature increases and the failure will be happened if
the tube could not sustain the changes.
2. MECHANICAL PROPERTIES
Hardness measurement has done using MicroVickers in NDT Lab PT Pupuk
Sriwidjaja for several samples. The data collections were being checked minimum 6
indentation for each locations.
The data for hardness for each samples are shown in table as follows,
NoHardness (HV)
A B C D E1 160 159 155 177 2362 161 160 150 163 2333 159 160 157 154 2534 163 155 168 152 2315 155 157 149 163 2356 155 157 150 168 2417 148 1878 153
Average158,833
3 158 156,6 171,6 238,1667Table 5. Hardness Measurement for each samples
As data shown above, there are increasing of average hardness trend when it come
to failure location. The closer to alligator lip, the harder material it be. This was
proved that deformation was taking role in failure, not creep or diffusion in phases.
Dislocations along grains were multiplied as time of failure reached the limit.
From Formula 3 and Formula 4, conversion was done to compare the tube standard
ASTM SA 106 Gr.B with recently tube condition. For tensile and yield strength,
standards for related material are,
Tensile Strength 415 Mpa (min)
Yield Strength 240 Mpa (min)
Table 6. Strength Properties of SA 106 Gr.B material
PAGE22OF31
And after the failure, the data for strength are shown in Table 6 as follow,
Location HV HBTensile Strength
(Mpa)
A 158,8333 159 564,45
B 158 158 560,9
C 156,6 157 557,35
D 171,6 172 610,6
E 238,1667 227 767,26
Table 7. Conversion of Hardness vs Tensile Strength for tube after failure in each location
The strength were increasing when it comes to fail. Before break, the yield and
ultimate point would be reached in maximum limit. The relationship between Ultimate
and Yield Strength with Temperature is shown in Figure 7 below. The higher
temperature of service, the weaker material to withstand stress. So, the standard of
451 Mpa for ulitimate tensile strength of SA106 Gr.B will be lower than initial.
Figure 7. Relationship between Ultimate and Yield Strength with Temperature
What we are concerning here is the recently strength and standard of ASTM 106
GrB which are all values in above limits. The sample C,D,E are above the standard
because of deformation was happened when failure and it proved about short-term
overheating. Samples A & B is the location where no failure effect was affected in
that area. But, the comparison between these values with standard which having 415
PAGE23OF31
MPa, it will be high concern for the next maintenance. The writers suggest taking
the sample for predicting what maintenance will be in the Auxiliary Boiler 101-BU
PUSRI-2 Plant.
3. PHYSICAL PROPERTIES
Another condition to prove short-term overheating failure that happened in tube is
thickening. Changes in tube ID and OD measurements can be indicators of
overheating. Increases of 5% or more are indicative of short-term overheating. The
measurement has been done in several location to classify degree of deformation.
REGION STANDARD (ASTM SA 106 GR.B) t(MM) % OF REDUCTION
1
5.08 mm (0,2”)
1,522,5
Average 2 60%
2
332,52,521
Average 2.33 54.1%
33,54
Average 3.75 26.2 %Table 8. Thickness Measurement for failure tube
Reduction of thickness in large number shown in Table 8 fully proven about the role
of deformation is more concerning in failure which can make cobra lip. This would
happen if material have lack of strength to withstand internal pressure in tube. The
thicker tube means the tube have lower strength to handle large deformation.
The diameter increase at the deformed area was approximately 5 times (± Ø 10 in.)
from the respective initial diameter. The OD tube surface around the bulge had
cracks (also referred to as mud flat cracking or elephant hide) and exfoliated oxide
patches that resulted from the oxidation and deformation that occurred during
PAGE24OF31
overheating as a result of thermal stressed caused by tube expansion and
contraction.
For the fluid quality, after investigation from the inner tube, the fireside was clean
means no additional sulfide, phosphate, etc was grown during the operation.
Viswanathan also said that the fluid was having less failure factor than debris or solid
blockaging media. So, it was learned from the plant investigations that there had
been no operating problems with the water softening system and that untreated
water had been used in the boiler.
The deformed area had thinned by 60% (Fig. 8) with respect to the original wall
thickness (0.508 cm, or 0.2in.). Boiler tube ruptures are typically labeled as thin-
lipped (which are related to short-term overheating) to describe the amount of
thinning (±60%) and diameter increase more than 10% at the bulged area. There are
features of short-term overheating. However, the no presence of any scale, mud
crack and oxide in the innerside reveal that longterm overheating was not present.
The thick tubes in some places and high tensile stress are two conditions which are
serious to handle. So, with the condition of recent tubes which still attached in
Auxilary Boiler until now, we are suggesting to handle the operation and
maintenance with care.
3mm
3mm
2.5mm
3mm
4mm
3.5mm
2.5mm
1.5mm
2mm
2.5mm
1mm
2mm
PAGE25OF31
Figure 8. The thickness measurement in failure tube
1 2 3
PAGE26OF31
4. METALLURGICAL EFFECTS
In the middle and last region of the boiler (at MK E), tube was finned to enhance
efficiency of heat transfer at this location. The tube had a 20 cm (7,8 in.) long crack
within a 8,5 cm (3,34 in.) deformed area ( Fig. 9.).
Figure 9. Cobra lip in Auxiliary Boiler 101BU Pusri 2 tube failure
The crack started at the OD of the tube. Note that, the crack is wide-mouthed and
burned area formed oxide (Fig 10)
Figure 10. The crack phenomenon which shown after failure
Oxide in fin
Initial crack
Transgranular cleavage
PAGE27OF31
Ahead of the crack tip, there are a number of voids that have formed at the grain
boundaries around the crack as a result of graphitization and grain boundary sliding
(Fig 11)
(a) (b)Figure 11. (a) Microstructure of crack lip ; (b) macrostructure of crack lip
A normalized microstructure of this type of tube contains approximately 60% pearlite
if the carbon content is at the high end of the specification. In this case, pearlite is
still visible near the tube OD and has deformed. Graphitization, in the early stages of
transformation, the carbide plates in the pearlite transform into spheroids. The
spheroids subsequently coalesce to form graphite. Graphite nodules were identified
in this study, and the only carbides found were mostly at the grain boundary (Fig
11.a). Decarburization due to short-term overheating is thought to be responsible for
the appearance of graphite on the finned area of the sample where the crack
occurred. The deformed grains and appearance of graphite observed in the
microstructure at the crack occurred and likely the usual grains at far of failure tube
noted the supportive of short-term overheating in the boiler tube.
graphite
Initial pearlite
Deformed pearlite
Deformed grain
250µm100x Magnification
PAGE28OF31
Figure 12. The place of SA 106 Gr.B in Fe-C Diagram
Creep failures are typically expected in superheaters and temperatures. However,
creep may occur under other conditions as well, provided that there are high firing
rates on the fire side (OD) along with the heavy scaling on the water side (ID), as in
the case, and a loss of strength due to decarburization. But in this case, if creep is
not presented means the awareness still in first concern due to degradation of
mechanical properties in some of tubes.
It was confirmed that part of the shutdown procedure for this boiler could result in
burner operation after draining water from the system. The authors believe that hot
gas from the burner impinging on the surface of the tube without cooling water cause
the short-term overheating.
SA 106 Gr.B
PAGE29OF31
5. THERMAL CRACKS
In August 13th,2010, after the failure happened, one crack tube was found in weld
joint tube-top Header and located at no.1 from East and no.4 from south and also
one leak was found in weld joint of bottom header and located at no.1 from East and
no.3 from South. Figure 13 below is shown the phenomenon,
Figure 13. Crack happened in weld joint of bottom header at no.1 from East and no.3 from South
Mechanism of above failure is thermal shock crack, which is originally a special type
of corrosion fatigue or stress-induced corrosion caused by alternating heavy heating
and cooling over a temperature span of some hundred degrees.
The mechanism of crack is explained as concerted action of environment and
mechanical stress (variations) leads to cracking. At thermal stresses, the material
makes the protecting magnetite layer break in narrow cracks in a more or less
parallel arranged pattern.
Figure 14. The mechanism of thermal crack
PAGE30OF31
H. CONCLUSION
The observed bulging and cracking of the tube was caused by SHORT-term
overheat followed by a brief episode of excessive overheating.
Internal blockage of tube, Loss of coolant circulation or low water level, Loss of
coolant due to an upstream tube failure and overfiring or uneven firing or boiler
fuel burners are the major cause of short-term overheating.
The heavy scale, was not developed in the inner tube, means it was pure of
short-term overheating
I. SUGGESTIONS
PAGE31OF31
I. BIBLIOGRAPHY
G.F. Vander Voort. (1986). Failure Analysis and Prevention,. Vol 11, Metals Handbook, 9th ed. American Society for Metals, Metals Park , 715–727.
Institute, A. P. (2003). API 573 : Inspection of Fired Boilers and Heaters .
Spry, J. P. (2005). An Overheat Boiler Tube Failure. ASM International .
Steam (38 ed.). (1972). New York: Babcock & Wilcox Co.