FAILURE OF SECONDARY

REFORMER DOME AND ITS REPAIR

A high pressure drop across the inlet and outlet of the secondary reformer in the Moranbah ammonia
plant suddenly occurred on Friday 10th May 2013. The plant was shut down to determine the cause.
After the catalyst and support alumina balls were removed it was found that parts of the refractory
support dome had collapsed and that refractory balls and catalyst had filled the bottom of the vessel
blocking the outlet transfer line.

Examination of the failed dome revealed that the bricks had primarily sheared at an angle close to 90
degrees to the major axis of the dome. It was concluded that the refractory dome failed as a result of
high thermal stresses introduced as a result of the different thermal expansion between the refractory
brick dome and the water cooled vessel shell. The design had no allowance for thermal expansion.

The dome was replaced with a novel design using readily available alumina blocks. The plant was
returned to service 30 days after the incident.

Nia Aliabadi, John McGrath, David Keen and Venkat Pattabathula

Incitec Pivot Ltd, Australia

David Firth
Quest Integrity, Wellington, New Zealand

This paper outlines the findings from the root

Introduction cause analysis and the innovative modifications
made to rapidly return the plant to operation.
A high pressure drop across the inlet and outlet
of the secondary reformer in the Moranbah Plant Background
ammonia plant suddenly occurred on Friday 10th
May 2013 when the plant was being restarted. The Moranbah ammonia plant designed by ICI
Excessive metal temperature (700oC (1300oF)) using their LCA process, produces 450 MTPD
was also noted in the refractory lined transfer of ammonia that is then converted to explosive
line. The plant tripped and was taken completely grade ammonium nitrate. The ammonia plant
off line to determine the cause. It was found that was originally first commissioned in Yazoo,
parts of the refractory support dome had USA in 1998 and operated there until 2003. It
collapsed and that refractory balls had filled the was then purchased by Incitec Pivot moved to
bottom of the vessel blocking the outlet transfer Moranbah and recommissioned in September
line. 2012.

2016 43
41 AMMONIA TECHNICAL MANUAL
 Moranbah in Central Queensland is roughly reformer is a heat exchanger type system where
1000km (600 mile) from Brisbane in a remote the exhaust gas from the secondary reformer is
location. It is strategically located in the heart of used to heat the outside of the catalyst tubes in
one of the world's largest coal mining provinces the primary reformer. However, the secondary
to supply explosives. reformer operates in an identical manner to one
in a typical ammonia plant. The secondary
Secondary reformer design, history and reformer details are given in Figure 1 and Table
normal operational conditions 1. The refractory design details are shown in
Figure 2.
The plants reformer/secondary reformer system
differs from a standard design as the primary

Figure 1. The Moranbah secondary reformer.

AMMONIA TECHNICAL MANUAL 44

42 2016
 Figure 2. Layout of refractory in the lower portion of the vessel

2016 45
43 AMMONIA TECHNICAL MANUAL
 Table 1. Secondary reformer details (dimension for lower catalyst section)

Internal Diameter 2883mm (9ft 5½ inch)

Wall thickness 76mm (3 inch) bottom
Shell material SA516 Gr 70 carbon steel
Refractory in shell 184mm (7¼ inch) of insulation + 127mm (5 inch) of hard face high alumina
refractory.
Water jacket From bottom dome up the shell fed by boiler feed water nominally at 100°C
(212°F)
Outlet temperature Normally 950 to 985°C (1742 to 1805°F). At failure the temperature rapidly rose
to 1080°C (1976°F). Note, in a normal ammonia plant the outlet is about 950°C.
Operating pressure 3400kPag (493psig)
Pressure drop Normally 115kPa (17psi)
At failure 900kPa (130psi)
Reformer outlet gas 11% CO, 6% CO2, 0.6% CH4, 30% N2 and 50%H2
composition
Dew point Approximately 177°C (350°F)
Support dome High alumina refractory bricks - Unicast 98, skew tiles - Greencast 95
material
Support dome layout 120 skew tiles supported on a steel shelf welded to the shell.
Single layer of tapered bricks in a dome with a 1981mm (6ft 6 inch) bottom
radius and 229 mm (9 inch) thick consisting of the followings rows of bricks .
• Row 0: 60 positioned next the skew bricks.
• Row 1: 50
• Row 2: 40.
• Row 3: 30
• Row 4: 14 row
• Central plug
Catalyst bed About 10 tonne of catalyst + alumina balls

The bottom of the row 0 bricks sit on the skew vessel design prior to recommissioning in
blocks (also called skew tiles) and are mounted August 2012.
at 35 degree to the horizontal. The skew blocks
are designed to fit against the inside of the After recommissioning in 2012, the Moranbah
vessel wall and sit on a steel support shelf. plant had numerous issues that resulted in the
plant having 42 restarts before the 10th May
The operational/history details available for the 2013 incident. These typically were unrelated to
secondary reformer prior to being installed at the reforming section of the plant and ranged
Moranbah were very limited. However, in 2002 from Electrical and Instrumentation faults to
at a turn around and catalyst change, the dome rotating equipment failure. Every stoppage of
was found to be damaged in some manner and the reformer area involves a rapid automated
was replaced. depressurization and 2 hour purge with nitrogen.
Typically this would reduce the dome
After the shipping to Moranbah, the second temperature to 650oC (1200oF)
dome was found to have collapsed. A new dome
was installed in accordance with the original

AMMONIA TECHNICAL MANUAL 46

44 2016
 Typically the secondary reformer operates at
steady conditions. During start up and shutdown
the secondary reformer is heated at a rate of
approximately 100°C/hour (180°F/hour) from
450 to 670°C (840 to 1240°F) then 230°C/hour
(414°F/hour) from 670 to 925°C (1240 to
1700°F) and cooled at a rate of 315°C/hour
(567°F/hour) from 980 to 750°C (1800 to
1380°F) then 25°C/hour (45°F/hour) from 750
to 550°C (1380 to 1000°F), see Figure 3. The
operational data was reviewed and there were
no trends present such as an increased pressure
drop that were observed.
Figure 4. Pressure drop increase from 20 to
900kPa and temperature during the start-up on
10th May 2013.

When the bottom manway, below the refractory

support dome, was opened it was found that the
dome had failed in a number of locations and
that alumina balls and catalyst had run into the
area below the dome and blocked the outlet, see
Figure 5. This also showed that the lower
portion of the majority of bricks in row 0 and
row 1 had broken off.

Figure 3. Normal temperature and pressure

drop profile during start-up of the secondary
reformer.

The incident on 10th May and observations

The plant had been running in a stable condition
from 5th of May when it tripped on 10th of May
at 2:11 am due to natural gas compressor
discharge temperature high high. At 9:30am
plant start-up commenced. At 17:09 pm after air
cutting to the secondary reformer, the pressure
drop suddenly increased from 104 to 825kPa
(15 to 120psi) in 8 mins and to 909kPag
(130psi) in 5 mins, see Figure 4. The plant
tripped during the start-up due to secondary Figure 5. Below the dome.
reformer outlet temperature high high (1080°C
(1950°F)).

2016 47
45 AMMONIA TECHNICAL MANUAL
 The catalyst and alumina balls were removed
from the vessel. Examination from above
revealed that bricks had completely failed and
were missing in three locations. These were in
the West, East (see Figure 6) and South East
Locations. In total 14 bricks in row 0, 7 in row
1, 2 in row 2 and 2 in row 3 had fallen out of the
dome after failing. The bricks from the dome
were removed. Examination of the skew blocks
revealed that a large number had cracked and
section had fallen off. After the removal of the
skew blocks no damage was found to the inside
of the shell or the steel support row which was
present below the skew blocks. The shell
diameter was measured and found to be within
±5mm (0.2in). Examination of the broken pieces Figure 6. The refractory dome from above
of the dome showed that the failure of all the showing the East failure – cracked skew blocks
bricks was consistent with the dome trying to and missing bricks from row 0, 1, 2 and 3.
expand and being restrained from expanding;
see Figure 7, i.e. the resultant force was radial.

Figure 7. Typical cross section of the dome after the failure aligned in comparison with marker to show
the horizontal plane.

The damage in the rows 0, 1 and 2 bricks is all High alumina bricks are hard and not readily
consistent shear loading across the bricks with crushed. Whenever they are designed to be used
the failures running near to parallel with the in the high temperature condition the design
horizontal. The damage to the skew bricks is should suitably allow for thermal expansion. For
consistent with high point loading where contact the design of the arch no thermal expansion had
had occurred with top edge of the row 0 bricks, been accommodated for in the design. Typically
i.e. the loading was no longer evenly spread this could include a softer mortar that is
across the face of the skew bricks. The centre crushable, expansion gaps or a flexible outer
plug failed as a result of the compression point portion such as alumina balls.
loading at the contact points with the tops of the
row 4 bricks.

AMMONIA TECHNICAL MANUAL 48

46 2016
 The refractory dome and the outer surface of the • Multi-layered brick arch. This structure is
steel will expand or contract by different far less stiff than a single layer of bricks and
amounts as a result of start-up/shutdown. For a allows for shear forces between the bricks.
length of 2883mm (9ft 5in), the design inside • A single solid plate of alumina mounted on a
diameter of the secondary reformer steel shell, it row a bricks. This design is similar to the
is calculated that this will increase in diameter pyramid cone as the expansions are taken up
by 3.5mm (0.14in) when it is heated from by movement of the alumina balls.
ambient temperature of 20°C (70°F) to water
jacket temperature of 120°C (248°F) assuming
an expansion rate of 12.5x10-6mm/mm/°C (7 There are very few complete failures of
x10-6in/in/°F). In comparison a high alumina secondary reformer domes that have been
refractory brick of the same length of 2883mm reported to have occurred. However, a failure
(9ft 5in) will increase in length by occurred in a single layer brick design dome in
22.5mm(0.9in) when it is heated from ambient 1982 at the Columbia Nitrogen fertilizer plant
temperature to 950°C (1740°F) (operating Augusta Georgia USA in an identical manner
temperature) assuming an expansion rate of [1] to the Moranbah event. However, the cause
(8x10-6mm/mm/°C (13 x10-6in/in/°F). The of the failure was not identified. In addition, it
difference in movement is 19mm. For a dome to was reported that damage occurred to the dome
be suitable for this application this movement of of the Moranbah vessel when it was operated in
19mm then needs to be compensated by a rise in Yazoo.
the height of the dome. However, the movement
is restricted by the weight and friction of the Incitec Pivot operates four secondary reformers
alumina balls/catalyst and the inability of the worldwide and this case was the first incident of
9in (229mm) thick rectangular bricks to rotate. a dome failure or any significant damage being
The design of the dome in a parallel seen to the domes.
(cylindrical) walled shell is not ideal as the shell
acts as solid immovable anchor. If the shell is Autothermal reformer after 4 years of
tapered at the location of the dome this may operation
allow movement up and down the taper.
Alternatively if the skew bricks Partial damage to refractory domes is not
had been at the more typical angle of 45 degree unknown as can be seen by this case study. The
the forces may have been lower. relevant data is shown in Table 2 and Figure 8.

There are a number of different designs used for

supporting the catalyst in secondary reformers.
These include:
• A pyramid cone of bricks with each row of
bricks sitting on top of the row below and
each row being in compression from the
weight of the catalyst supported on alumina
balls. This design allows for thermal
expansion by movement in the alumina
balls.

Figure 8. ATR dome design

2016 49
47 AMMONIA TECHNICAL MANUAL
 Table 2. ATR
Age 4 years
Shell diameter 2.25m (7ft) at base of dome
Outlet temperature 950°C to1050°C
Shell temperature Water cooled at about 100°C (212°F)
Support dome Single layer of 300mm (12in) high alumina
Z bricks with dome radius of 1700mm (5ft
6in). The centre plug was made from 4
quadrant bricks
Load supported by dome Approximately 20 tonnes static load and 2
barg differential pressure dynamic load

The Incitec secondary reformer at Gibson Island

A shutdown inspection revealed that all four of (GI) Brisbane Australia has a catalyst support
the quadrant centre bricks had fractured structure made from alumina blocks spaced in
horizontally across the Z step line. The lower lattice to allow gas flow. This had been in
halves were lying on the floor below. The service for 10 years and no major issues had
catalyst was removed so that the dome could be been seen.
inspected. Note, health and safety requirements
forbid the entry under the dome with the catalyst A number of alternative designs were
in place. No significant damage was found to considered at Moranbah. It was decided to
any of the other bricks. However, it was found create a structure out of alumina blocks similar
that the base of the dome was 25mm (1in) off to that at GI using readily available material.
being horizontal due to a variation in the The design used is shown in Figure 9.
fabrication of the steel shell and the original
installation. The fractured centre plug bricks
were removed and a new insert was cast in
place.

Returning the Moranbah secondary reformer

to service
When the incident occurred at Moranbah the
plant had suffered numerous issues and a spare
refractory dome was not available. For a dome
the bricks designs vary between each row and as
result it is not possible to get delivery of high
alumina bricks urgently off the shelf. In
addition, it was considered highly risky
replacing the dome with the same design
considering the history of the vessel. After the
failure in Augusta they used a “goal” post Figure 9. The new design for catalyst support
design in front of the outlet. This was made for the Moranbah secondary reformer.
from an Inconel 800 frame.
The new installed dome and repaired refractory
required a very delicate dry out process this

AMMONIA TECHNICAL MANUAL 50

48 2016
 included four major steps as shown in Figure plant was returned to service after being
10. shutdown for 30 days.

The catalyst support structure was installed, new References

catalyst was used and the plant was returned to
service 30 days after the incident. Since
installation the pressure drop has been within 1. George E. Connaughton and Robert W.
normal design window ranging from 101 to Clark Columbia Nitrogen Corp. Secondary
108kPa (14.6 to 15.6psi). Reformer Catalyst Support Dome: Failure and
Repair, AICHE 1983.

Figure 10. Refractory dry out and initial start-

up of the secondary reformer after the
installation of the new catalyst support structure
over a five day period.

The support structure was inspected in March

2014. Upon inspection by a refractory expert it
was seen that the refractory was in serviceable
condition with no areas of spalling or significant
cracking. Minor repairs were required to the
refractory on the associated transfer piping to
remove minor hotspots.

Conclusions
The Moranbah catalyst support dome failed as a
result of high thermal stresses introduced as a
result of the different thermal expansion
between the refractory brick dome and the water
cooled vessel shell. The design had no
allowance for thermal expansion and high
number of thermal cycles.

The dome was replaced with a novel design

using readily available alumina blocks and the

2016 51
49 AMMONIA TECHNICAL MANUAL