EXPERIENCE ON FAILURE AND REPAIR OF AMMONIA

SYNTHESIS CONVERTER DUE TO HYDROGEN CRACKING

Farabirazy Albiruni, Wildan Hamdani, Posma M Pakpahan

Technical Bureau of Inspection, PT. Pupuk Kalimantan Timur, Tbk.,
Jl. Ir. James Simanjuntak No. 1, Bontang 75313, East Kalimantan, Indonesia
Email: [email protected], [email protected], [email protected]

Abstract

Ammonia converter is a pressure vessel that provides conversion of synthesis

gas (70 % wt. of hydrogen gas) into ammonia. In January 2005, a leaky
transverse crack was detected in circumference weld 2 (CW2) of Kaltim 4’s
ammonia converter (1-R-501) after operating for two years. Failure analysis
then performed to find the cause(s) of the failure. It consists of visual
observation, positive material identification, penetrant testing, conventional
ultrasonic testing, hardness testing and metallographic observation. These
results showed that this failure suspected by hydrogen cracking due to local
improper post weld heat treatment (PWHT). Improper PWHT can introduce
high residual stresses. This vessel made by 2.25 Cr-1 Mo steel and very
prone to form martensite structure during welding.

The repair procedures consist of hydrogen removal followed by conventional

ultrasonic test, re-weld the failure location in two stages with each stage was
followed by hardness testing, magnetic particle testing, high temperature
penetrant testing, and time flight of diffraction (TOFD).

After one year since the repair time, a high temperature conventional
ultrasonic testing was performed to this repairing circumference weld and
showed good result. It is then recommended to perform conventional
ultrasonic test or TOFD annually to assess the condition of the vessel.

Keywords: Ammonia Synthesis Converter, Hydrogen Cracking,

Improper Post Weld Heat Treatment, 2 ¼ Cr–1 Mo Steel,
Conventional High Temperature Ultrasonic Testing, Time of
Flight Diffraction (TOFD).

1. INTRODUCTION inspection. Table 1 below gives the

vessels’s technical specification.
Ammonia converter is a pressure
vessel that provides conversion of
synthesis gas into ammonia. On 2. PRE – REPAIR INSPECTION
January 2005, there was a leak on shell
of Kaltim 4’s ammonia converter (1-R- In order to know the cause(s) of
501). This leaky indication found by the failure and vessel’s condition, the
regular explosive gas level inspection then performed and it
measurement that reached 100% consists of:
explosive level at top cover. Verification
using Infra Red ThermaCam (Figure 1) 1. Visual Observation
showed a hot spot indication around the 2. Positive Material Identification
south side thermowell nozzle (180o 3. Liquid Penetrant Testing
orientation; North = 0 o). The adjacent 4. Conventional Ultrasonic Testing
temperature to hot spot area was 25 oC 5. Hardness Testing
while on it was 62 oC. The ammonia
plant had have to shut down then for
2. 1. Visual Observation 2. 2. Positive Material Identification

Visual observation found a leaky This test performed on weld and

transverse crack (40 mm in length) on base metal of CW2. The result is shown
circumference weld 2 (CW 2) as shown in Table 2 below. It is obvious that the
in Figure 2 below. The crack position weld and base metal are made from 2 ¼
was 100 mm in distance after south side Cr – 1 Mo steel. This result agreed with
thermowell nozzle (180o orientation). the vessel’s material specification.
This result agreed with the result from
ThermaCam observation and explosive 2. 3. Liquid Penetrant Testing
gas level measurement that showed the
area around the south side thermowell Liquid penetrant testing performed
nozzle suffered leakage. on CW 2 and longitudinal weld 1 (LW1).
The visual observation also There was one crack indication found on
showed that the color of kao wool heat affected zone (HAZ) of CW2 at
insulation around the leakage zone had 100o orientation with 20 mm in length, as
changed from white to brown compare shown in Figure 3 below. This indication
to non leakage zone. dissapeared after grinding up to 4 mm
depth from surface.

111.8°C

100

80

60

40

Hot Spot 20
Area
6.7°C

Figure 1. The ThermaCam View of Hot Spot Indication

Table 1. Technical Specification of Kaltim 4’s Ammonia Converter (1-R-501)

1 – R – 501
Design Code : A. D. Merkblatter
Description : Ammonia Synthesis Converter
Fluid Contained : Synthetis Gas
Design Temperatute (oC) : 370
Operating Temperature (oC) : 270
2
Design Pressure (kg/cm G) : 155
Operating Pressure (kg/cm2 G) : 132
Shell Material : SA 542 B Cl. 4
Thickness of Shell (mm) : 105.0
Corrosion Allowance (mm) : 1.70
Years in Service : 2003
Composition of Synthetic Gas (%wt.) : H2 = 71.41, N2 = 27.5, CH4 = 0.76
 Table 2. Positive Material Identification Result
Composition (% Wt)
Location
Cr Mo
Weld Deposit 2.21 1.02
Upper Base Metal 2.27 0.99
Lower Base Metal 2.23 1.12

CW-2

N
( 0o )

100

( 180o )
S

Figure 2. A Leaky Transverse Crack on

Circumference Weld 2
 Surface
Crack

Figure 3. A Surface Crack Found by Penetrant Testing

2. 4. Hardness Testing

Hardness testing performed both LW1. These results are given in Table 3
on base and weld metal of CW2 and to Table 6 below.

Table 3. Hardness Test Result on CW2

North
1

15 5

10
Heat Affected Zone Base Metal
Weld Deposit
No Upper Lower Upper Lower
( BHN )
( BHN ) ( BHN ) ( BHN ) ( BHN )
1 178 165 204 145 143
2 168 171 141 138 142
3 171 168 144 136 165
4 162 158 166 140 137
5 158 171 140 133 151
6 177 163 148 140 158
7 192 171 162 140 170
8 171 158 180 169 166
9 192 153 168 140 175
10 240 153 162 141 174
11 260 160 180 162 148
12 247 169 216 165 165
 Heat Affected Zone Base Metal
Weld Deposit
No Upper Lower Upper Lower
( BHN )
( BHN ) ( BHN ) ( BHN ) ( BHN )
13 275 162 151 163 160
14 195 171 165 155 165
15 245 172 170 153 171
16 252 178 165 161 158
17 260 151 147 147 140
18 174 151 157 140 140
19 180 153 160 153 148
Note: Distance between position of 10 to 13 = 1140 mm
Distance between position of 15 to 17 = 850 mm

Table 4. Hardness Test Result Around Leakage Zone ( 5 – 20 mm from Leakage )

5 mm

I II III IV V VI VII VIII

1.
2.

3.
4.
5. Weld Deposit
6.

7.
8.
Leakage

Weld Deposit HAZ Base Metal

No (BHN) (BHN) (BHN)
3 4 5 6 2 7 1 8
I 287 290 257 287 192 185 147 172
II 290 283 259 280 187 187 161 171
III 271 290 271 249 179 188 153 162
IV 280 274 265 255 255 228 156 148
V 288 282 295 233 227 197 158 145
VI 269 241 255 282 211 177 147 172
VII 253 295 284 295 187 185 159 171
VIII 261 287 263 295 255 210 147 140

Table 5. Hardness Test Result on Longitudinal Weld 1 (LW1)

Heat Affected Zone Base Metal
Weld Deposit
No Left Right Left Right
( BHN )
( BHN ) ( BHN ) ( BHN ) ( BHN )
1 185 177 194 162 162
2 185 185 194 154 162
3 177 194 185 140 162
4 185 177 194 147 145
5 194 186 194 162 154
6 192 194 200 147 154
7 187 220 194 147 154
8 174 175 177 140 154
Note: Measurement Performed from Top to Down
 Table 6. Hardness Test Result Around Surface Crack (5-20 mm from Crack)
5 mm

I II II III IV V VI

1.
2.

3.
4.
5. Weld Deposit
6.
Surface Crack
7.
8.

Weld Deposit HAZ Base Metal

No (BHN) (BHN) (BHN)
3 4 5 6 2 7 1 8
I 164 165 177 171 192 186 150 141
II 170 163 160 192 187 185 140 140
III 171 156 153 150 179 177 138 148
IV 171 164 158 158 181 179 145 140
V 160 160 160 165 180 177 148 140
VI 160 163 157 160 179 181 143 146

2. 5. Conventional Ultrasonic Testing linear indication as shown in Figure 4

and Table 7 below. This linear indication
Conventional ultrasonic testing had 21% level of Distance Amplitude
performed on CW2. This test found one Correction (DAC).

Leakage

Figure 4. Indication Found by Conventional Ultrasonic Testing on CW2

Table 7. Detail of Linear Indication Found by Conventional UT

Position Depth
No Width (mm) Remark
(o ) (mm)
1 30 10 43 Linier Indication, 21% DAC
3. PROCEDURE OF REPAIR and the rest must be cut out using
pen grindstone until reach the inner
The procedure of repair consists surface of shell. Every reach the 10
17 stages as follow: mm depth, the inspection on crack
performed. Keep the constant
1. Hydrogen removal on CW2: leakage zone temperature at 200 oC
during gouging and grinding aided
A. Attach one layer of heating by using oxy-acetylene torch. Clean
element on CW2. the gouging by grindstone and also
B. Increase the temperature up to for bevel finishing. Penetrant testing
400 oC ± 5 oC with rate of 50 performed to assure there was no
o
C/hour. crack left on the shell. Figure 5
C. After reach 400 oC, hold the below shows the result of gouging
temperature for 5 hours. and grinding the leakage zone
D. Decrease the temperature up to meanwhile Figure 6 shows the
200 oC with rate of 50 oC/hour. penetrant testing result performed
E. After reach 200 oC, open after.
insulation and unattach the
heating element. 4. Attach the heating elements on
leakage zone. Figure 7 shows
2. Perform conventional ultrasonic schematically the arrangement of
testing on CW2. heating elements placed on leakage
zone
3. Gouge the leakage zone out to 95
mm depth in order to form the bevel

Crack/ Leakage 200 mm

75 mm

Figure 5. The Leakage Zone Condition During and After Gouging

Crack Rest

60 mm
Height of the Hole
After Grinding

Figure 6. Penetrant Testing Result Performed After.

 Attachment of
two heating
elements
Area equal to one
heating element

TC WELD

Bevel

Figure 7. Heating Elements Arrangement

5. Increase the heating elements 7. After welding up to 50 mm finish,

temperature up to 225 oC ± 10 oC for attach heating elements with two
weld preheating and keep it constant layers arrangement and
during welding process. thermocouple on welding zone for
local Post Weld Heat Treatment
6. Root pass welding up to 50 mm (PWHT). On shell surface there was
height of weld deposit using SMAW. an incoloy plate to avoid heating
During root pass welding, turned off elements overheating. Figure 8
the nitrogen gas purging inside shows schematically the process of
vessel for a while. PWHT

Heating 2 Incoloy Plate

elements

3 Monitoring
4 Thermocouple
Controlling
Thermocouple

50 mm Weld
Deposit

Figure 8. Local PWHT Process

8. Perform PWHT process as follow: B. After reach 450 oC, hold the
temperature for 4 hours.
A. Increase the temperature up to C. Increse the temperature up to
450 oC ± 10 oC with rate 50 720 oC ± 10 oC with rate 50
o o
C/hour. C/hour,
 D. After reach 720 oC ± 10 oC, hold example, during holding at
the temperature for 3 hours. 720 oC, thermocouple (T/C) 1
E. Decrease the temperature up to temperture was 700 oC, T/C
200 oC with rate 50 oC/hour. 2 = 800 oC, T/C 3 = T/C 4 =
F. After reach 200 oC, turn off 720 oC.
PWHT machine and unattach
the heating elements. 9. Grind the surface of weld deposit to
attain the smooth surface. Figure 9
Note: During PWHT, there was a shows the weld deposit surface
difficulty to attain the uniform condition after grinding.
temperature distribution. For

Figure 9. The 50 mm Weld Deposit Surface after Grinding

10. Perform non destructive testings 15. Attach the heating elements and
consist of hardness testing, thermocouple for this repair weld
magnetic particle testing, high joint.
temperature penetrant testing and
time of flight diffraction (TOFD). 16. Perform final post weld heat
TOFD performed on whole weld of treatment (PWHT) on whole CW 2
CW2 after the shell temperature as follow:
reached 50 oC.
A. Increase the temperature with
11. After NDTs give the good results, rate 50 oC/hour up to 740 oC±10
o
reattach heating elements C (for weld deposit) and 650
o
circumferencially on CW2 except on C±10 oC (for base metal).
leakage zone. The arrangement of B. Hold the temperature for 4 hours.
heating elements same as shown in C. Decrease the temperature up to
Figure 7. The heating elements 200 oC with rate 50 oC/hour.
attached on weld joint consist of two D. Turn off the PWHT machine and
layers. unattach the heating elements.

12. Increase temperature up to 250 oC ± 17. Perform non destructive testings

10 oC with rate 50 oC/hour for weld consist of hardness testing, high
preheating and hold during welding temperature penetrant testing and
process. time of flight diffraction (TOFD).
TOFD performed on whole weld of
13. Continue welding process to full CW2 after the shell temperature
thickess of shell. reached 50 oC.

14. Grind the surface of CW2 for

hardness testing.
4. ON AND POST – REPAIR 4. 1. Hardness Testing After Weld Up
INSPECTION to 50 mm of Thickness

Non destructive testing such as A. Repair Area

hardness testing, high temperature Table 8 shows hardness testing
penetrant testing, magnetic particle result on repair area.
testing, conventional ultrasonic testing,
and TOFD were applied during and post B. Non Repair Area
repair inspection. Table 9 shows hardness testing
result on non repair area.

Table 8. Hardness Test Result on Repair Area

4
11 · 5
· ·
·
1

10 Weld
2 6
Deposit
· · ·
9 3 7
· 8 · ·
·

Hardness (BHN)
Position
1 2 3 4 5 6 7 8 9 10 11
189 162 162 - 142 148 140 168 162 187 168
168 157 147 - 143 158 142 177 154 177 175
163 160 150 - 140 150 145 156 151 201 174

Table 9. Hardness Testing Result on Non Repair Area

270 mm

200 mm 1 2 3 4 5 6

Heating
element Place 7 8 9 10 11 12

13 14 15 16 17 18 19

20 21
TC 22 23 24 25 Weld Deposit

26 27 31 32
28 29 30

Repair Area
33 34 35 36 37 38
 Base Metal Base Metal Weld Deposit
Position Position Position
(BHN) (BHN) (BHN)
1 175 163 156 17 145 147 147 20 280 277 282
2 147 166 148 18 145 147 145 21 275 267 270
3 161 155 153 19 156 161 170 22 192 188 202
4 157 158 153 26 182 195 194 23 177 180 187
5 151 158 160 27 177 195 171 24 206 211 204
6 147 150 160 28 143 140 145 25 189 191 189
7 175 156 157 29 143 145 140
8 160 153 169 30 140 145 150
9 171 154 165 31 140 140 147
10 165 158 166 32 160 167 165
11 145 148 148 33 160 161 174
12 153 160 155 34 167 170 175
13 153 145 150 35 164 174 161
14 145 147 147 36 170 171 175
15 145 150 157 37 174 167 172
16 171 153 160 38 151 164 158

4. 2. Hardness Testing After Final

PWHT
.
Hardness testing result after final
PWHT is shown in Table 10 and 11
below.

Table 10. Hardness Testing Result after Final PWHT

North
1

15 5

10
Heat Affected Zone Base Metal
Weld Deposit
Position Upper Lower Upper Lower
( BHN )
( BHN ) ( BHN ) ( BHN ) ( BHN )
1 154 165 188 140 140
2 165 168 140 140 142
3 171 166 155 140 155
4 160 165 165 140 138
5 158 177 140 130 148
6 165 165 152 140 140
7 150 177 160 135 175
8 130 155 162 166 155
9 120 160 145 138 171
10 133 150 165 138 168
11 180 162 171 160 146
12 171 165 182 162 160
13 175 165 145 160 158
14 175 170 166 146 145
15 175 168 166 154 170
16 180 177 165 166 155
17 165 145 154 151 140
18 165 145 154 141 140
19 165 151 171 150 152
Note : Distance between position 8 to 10 is = 540 mm
 5 mm Table 11. Hardness Testing Result around Repair Weld

I II III IV V VI VII VIII

1.
2.

3.
4.
Weld Deposit
5.
6.

7. Repair Weld
8.
Weld Deposit (BHN) HAZ (BHN) Base Metal (BHN)
No
3 4 5 6 2 7 1 8
I 185 182 179 170 177 169 138 169
II 172 179 192 175 154 162 138 162
III 177 165 169 171 154 171 147 162
IV 163 145 151 165 140 162 138 162
Repair
126 132 140 162 - - - -
Weld
V 121 140 123 125 140 128 120 140
VI 122 120 120 120 124 125 120 120
VII 129 133 140 138 126 133 120 122
VIII 137 132 133 135 137 133 140 154

4. 3. Penetrant Testing, Magnetic of these applied non destructive

Particle Testing, Conventional testings.
UT, and TOFD The indications found by
conventional UT and TOFD were
During and post repair, some of accepted indications and no need to
non destructive testings were repair. Figure 10 and 11 show the TOFD
performed. Table 12 shows the results results on repair area.

Table 12. On and Post Repair Non Destructive Testings

Test Result
Repair Stage
Penetrant MPI Con. UT TOFD
After Hydrogen Removal --- --- Acc ---
After Local PWHT (welded
Acc Acc --- Acc
up to 50 mm)
After Final PWHT Acc --- Acc Acc

Outer surface

Depression

Indications
Inner surface

Figure 10. TOFD Result on Repair Area (Welded up to 50 mm)

 Figure 11. TOFD Result on Repaired Area (After Final PWHT)

4. 4. High Temperature Conventional during testing. The accuracy of the

UT defect’s depth measurement is likely
within ± 10% accurate meanwhile the
High temperature conventional height of defect is greater than actual
ultrasonic testing performed after one (could be up to 100 % more than
year since the repair time. This testing actual). From this testing known that all
conducted on vessel during its operation indications found are acceptable and
with outer surface temperature reach showed that vessel in good condition.
240 oC. The gain setting for this testing Table 13 shows the result of this test
was 80-85 dB. In this gain range, this while Figure 12 shows the result on
testing is not suitable for detection of repair area schematically.
defects that may occur within the outer
0-25 mm range due to high level noise

Figure 12. High Temperature UT Indications on Repair Weld

Table. 13. High Temperature Conventional Ultrasonic Testing Result

Indication No. Depth (mm) Height (mm)
#1 50 2
#2 60 4
#3 75 10
#4 24 3
#5 ID Surface <3
#6 ID Surface <6
5. DISCUSSION expansion coefficient at this temperature
of 15 x 10 -6 mm/mm/oC. Under this
Ammonia synthesis converter of condition creep will readily occur and the
Kaltim 4 (1-R-501) had failed after 2 stress will recur. If this temperature
years in operation. The failure occured variation is such that the inside surface
was a leaky transverse crack on of the vessel is hotter than the outside
circumference weld 2 (CW2) which had one at temperature it will result in high
applied local PWHT during vessel residual tensile stresses on the inside.
fabrication. This may often be the case for
Hardness test from outer shell circumference welds, which are allowed
surface shown that CW2 around the to cool from the outside of the vessel.
leakage zone have hardness value Heseur et al (5) modelled this effect and
above the maximum value after PWHT, indicated that yield magnitude residual
that is more than 225 Brinel Hardness stress can be formed as a result.
Number (BHN) according to NACE The synthetis gas contained in this
standard MR0175 and API 942 (6,7). On vessel is rich with hydrogen gas (± 70%
other area of CW2 far from leakage wt H2 gas) with operating temperature
zone also found the hardness value around 300 oC. With improper local
more than 225 BHN. This is most PWHT applied, not all martensite
probably caused by improper local transformed into tempered martensite
PWHT during fabrication, such as: and residual stress reduce significantly
1. Local PWHT using heating elements as shown by hardness results. This
only applied on one side of shell makes the vessel still has high
surface. sensitivity to high temperature hydrogen
2. Lack of thermocouple amount to attack and failure occured in two years
control uniformity of temperature. of operation of this vessel.
3. Insufficient temperature and or During repair, re-PWHT had
holding time during PWHT. applied on CW2 in order to decrease the
The 2 ¼ Cr – 1Mo steel as a shell sensitivity to high temperature hydrogen
material has greater tendency to form attack by transformed all retained
martensitic structure during welding martensitic structure into tempered
because the shift of martensite martensite and also decrese the residual
equilibrium line influenced by Cr and Mo stress. For leakage zone, two times
(Cr and Mo will influence the Carbon PWHT had applied, those are after weld
Equivalence Number). Martensite has up to 50 mm and after weld into full
brittle and hard characteristic and very thickness 110 mm. The PWHT
prone to high temperature hydrogen temperature 740 oC ± 10 oC higher than
attack. This is why welding of 2 ¼ Cr – ASME recommendation 677 oC (8). This
1Mo stell must be followed by proper is because from simulation using carbon
PWHT in order to get more ductile steel plate 100 mm thickness found the
martensitic structure known as tempered different temperature of 100 oC – 120
o
martensite. Beside that, in thick walled C between two contact and non contact
vessel, application of PWHT is intended surface with heating elements. Because
to reduce the residual stress introduced the repair only attach the heating
into material during welding. This elements from outer shell surface,
residual stress can be as high as the higher temperature with two stages
yield stress near the inner surface of the PWHT intended to reach the inner
shell after PWHT if a temperature surface of shell because this surface
gradient occurs accross the wall during has direct contact with syntetic gas and
PWHT. For example, in a restrained 100 decrease its sensitivity to high
mm thick wall echanger, the 2 ¼ Cr – 1 temperature hydrogen attack.
Mo steel will start to creep (100 Mpa at The hardness test after final
700 oC) with a temperature variation of PWHT show some area with relatively
30 oC across the wall will have a surface low hardness of 120 – 130 BHN caused
stress of 100 Mpa (assuming thermal by higher PWHT temperature. This is
only happen on the outer surface and 6. ____., ”Petroleum and Natural
not in the inner surface because the Gas Industries—Materials for
temperature gradien during PWHT. Use in H2S-containing
Environment in Oil and Gas
6. CONCLUSION Production Parts 1, 2, and 3”,
NACE MR0175 / ISO 15156-
The failure of this vessel weld joint 1:2003 (E).
suspected by hydrogen cracking caused 7. ____., ”Controlling Weld
by improper local PWHT applied during Hardness of Carbon Steel
vessel fabrication. Improper PWHT can Refinery Equipment to Prevent
introduce high residual stress into Environmental Cracking”, API
material. Besided that, the 2 ¼ Cr – 1 942.
Mo steel is very prone to form 8. ____., ”Rules for Construction of
martensitic structure during welding. Pressure Vessels”., ASME Sect.
With improper local PWHT, not all VIII Divison 1&2., 1998.
martensitic structure transformed into
tempered martensite and residual stress
reduce significantly. This is why the 9. BIOGRPHIES
vessel had failed in very short operating
time. Farabirazy Albiruni
obtained his B.Eng.
7. RECOMMENDATION degree from University of
Indonesia in Metallurgy
Performed Conventional UT or and Material Engineering.
TOFD annually to assess the current He is currently working as
vessel condition. a Metallurgist and NDT
Engineer in PT. Pupuk Kalimantan
8. REFERENCES Timur, Tbk.

1. Timmins, P. F., ”Solution to Wildan Hamdani is a

Hydrogen Attack in Steels”.,ASM senior inspection engineer
International., USA., 1997. at PT. Pupuk Kalimantan
2. Prescott, G. R., Shannon, Brian., Timur, Tbk. He holds
”Process Equipment Problems B.Eng degree in
Caused by Intercation with Mechanical Engineering
Hydrogen”., Ammonia Plant from Diponegoro
Safety & Related Facilities p.p University. He is currently pursuing his
237-252., USA., 2001. Master Degree at Gadjah Mada
3. Firth, D. M., Keen, D., Jones, C., University in Maintenance Engineering
Kartensen, A., ”Cracking and field.
Repair of Closing Welds in 2.25
Cr 1 Mo Steel Vessels Operating Posma M Pakpahan is
in High Temperature Synthesis an Inspection Manager at
Gas”., Ammonia Plant Safety & PT. Pupuk Kalimantan
Related Facilities pp. 161-169., Timur, Tbk. He holds
USA., 2005. B.Eng degree in
4. Prescott, G. R., ”Weld Failure in Metallurgical Engineering
a 2 ¼ Cr – 1 Mo Ammonia and Master degree in
Converter”., Ammonia Plant Material Science both from University of
Safety & Related Facilities pp. Indonesia.
217-231., USA., 1992.
5. Heuser, Albert., ”Repair of a
Thick-walled Ammonia Synthesis
Converter Containing a Leak”.,
AIChe., USA., 1991.