Date: 27/11/2012
STRIPPER ACCIDENT AT
THE UREA FERTILIZER
FACTORY LTD,
GHORASHAL, BANGLADESH
Syeda Sultana Razia, PhD
Dept of Chemical Engineering, BUET
Case Study: Carbon Dioxide Stripper Failure
An accident occurred with the stripper in the
renovated 1,422 tons/day Urea Fertilizer Factory
Limited (UFFL), located in Ghorashal, Bangladesh
at 00:10 am on June 20, 1991
11 employees of UFFL died, several others
including employees of Toyo Engineering
Corporation (TEC), Japan, (the General
Contractor)were injured
Material loss: Estimated USD 62.5 million
References
Report of the Enquiry Committee on Ghorashal
Fertilizer Factory Accident, Ministry of Industries,
Gov’t of Bangladesh, July, 1991
Investigation Report for Stripper accident:
Ghorashal Fertilizer Renovation, UFFL,
Bangladesh, Toyo Engineering Corporation, Tokyo,
Japan, August, 1991
Background
The existing 660MT/day Ammonia and 1137 MT/ day
Urea plants were renovated to achieve 25% increase
of design production capacity and improvement of
energy efficiency
The existing synthesis process was converted to
energy saving synthesis process called ACES
(Advanced Process for Cost and Energy Savings)
New synthetic section consisted of five high pressure
equipment including the urea reactor, CO2 stripper,
two carbamate condensers and scrubber
Chemistry of Urea Production
Raw Material : ammonia and carbon dioxide
Exothermic reaction of NH3 and CO2 to produce
ammonium carbamate:
2 NH3 + CO2 ↔ H2N-COONH4
Endothermic decomposition of ammonium
carbamate into urea and water:
H2N-COONH4 ↔ (NH2)2CO + H2O
Urea
Block Diagram of Synthesis Section
Scrubber
CO2 stripper
CO2
PFD of Synthesis Section
Stripper Description
The stripper: shell and tube heat exchanger in
vertical position
Tube side (feed):
Top: reactor effluent (urea products and carbamate)
Bottom: CO2
Shell side: Steam
ID of the shell: 2.3m
Height of the shell 13.3m
Pressure vessel design code: AD Merkblatter
(Germany)
Material : special duplex stainless steel (Lining
of channel:25 Cr-6Ni-1.5 Mo and Tube: 25 Cr-7
Ni- 3Mo)
Stripper Design and Operating Conditions
Shell Side Tube side
Design Pressure 26 184
(kg/cm2G)
Design Temperature 240 220
(0C)
Operating Pressure 20 175
Operating 214 190/176
Temperature
Corrosion allowance 3 Lining thk
(mm)
Hydrostatic Test 33.8 239.2
Pressure(kg/cm2G)
Non-destructive inspection: PT (liquid penetrant examination) and
RT (radiographic examination of dissimilar metal weld joints) were
Incident Description
The plant was operated according to the start-up
plan. CO2 gas was flowing through the stripper.
Liquid ammonia feed pump and recycle solution
pump were started. The pressure of the ACES loop
reached 160 kg/cm2G
There were two loud explosion sounds. The high
pressure stripper leaked to atmosphere. The stripper
bottom fell out and the main body of the stripper was
lifted up 30-35 ft and fell on the ground penetrated
14ft into the ground
Position of the Stripper
Photographs
CO2 Inlet
Distributor
CO2 Inlet Distributor Movement at Accident
Consequences
The severed pipes released ammonia, CO2 gases
and urea and carbamate solutions
The power in the control and adjoining areas went
off
About 50 people in and near the control room were
affected and 7 (of total 11) died on spot
Control Room
21 m away from the
stripper
Was on the second
floor
With single exit stair
What Caused the Accident?
No prior warning “Overpressure or Explosion”
No indication of malfunction from control panel
Nor were any visual observations
Possibility of Overpressure?
All indicators installed in the synthesis section
showed pressure lower than the design pressure
Nobody recognized any alarm or sound of safety
valve discharge
Pressure Profile of the Stripper
Design Pressure
Possibility of Explosion?
Flammable gases are H2 and/or NH3
Inert gas are CO2 and/or H2O
With fully open air feed valve to compressor: air
content is 15 vol%
Hydrogen content is at most 0.7 vol%
Back flow of ammonia gas from reactor through
carbamate condenser
Flammable Ranges
CO2 H2O
H2
Nonflammable NH3 Nonflammable
Flammable
Flammable
Air Air
For Hydrogen For Ammonia
Flammable Ranges
For NH3 and H2 mixture: using Le Chatelier rule
CO2+H2O
NH3+H2
Air
Gas Composition Ranges
CO2 charging NH3 mixing
Temperature Profile of Stripper
Design
Temperature
The Accident was not caused by
Overpressure or Explosion
Detailed Inspection
No burn or rise of
temperature on plastic
gaskets of internals
No change of metallic
color of internals
Fractured surface of both
tubsheet and bottom head
shows a black zone at the
4th layer between the
lining and outer surface
Crack initiated at this
zone and propagated to
inner and outer direction
Tests/Inspection of Fractured Stripper Samples
Chemical analysis: fluorescent x-ray analysis
and vacuum emission spectro-chemical
Met the code requirement
H2 content of weld metal: 6ppm ( outer layer) and
up to 10ppm (inner layer)
Mechanical properties: Tensile strength and
Charpy Impact strength
Met the code requirements
Nondestructive inspection: RT, UT and PT
No defect was detected
Tests/Inspection of Fractured Stripper Samples
(continued)
Microstructure and hardness: EDS ( Energy dispersive
spectrometry) and (X-ray diffraction)
First layer of saw-wire buttering weld is transformed to
martensite
Black zone consisted of magnetite: formed by high
temperature oxidation, hence, the welding crack existed
before the vessel fractured
Fractrographic features with SEM:
Fracture surface is in intergranular fracture mode with
innermost surface ductile : probable hydrogen embrittlement
Test Results
The welding of dissimilar metals i.e. tubsheet
surface and seam of lining of channel inside was
not done in accordance with Welding Procedure
Specification (WPS): welding crack existed along
the deposit metal
Not detected by radiographic examination at the shop
Not ruptured the stripper during hydrostatic test
The most probable cause of rupture was hydrogen
embrittlement ( requires 6 to 10 ppm H2)
Investigation of the Enquiry Committee
UFFL was constructed by Toyo Engineering
Corporation (TEC) of Japan on a turn key basis on
May 1970 funded by Japanese Yen credit
Implementation of Rehabilitation scheme by TEC in
1985 improved the performance UFFL significantly
The same year TEC proposed yet another renovation
of UFFL with newly developed ACES process with a
favorable finance from Japanese Govt - this
renovation was not required or asked by BCIC
Findings and Responsibility
TEC only had experience of new ACES process for
smaller plants and scale-up of 200% was required for
UFFL
CO2-stripper and carbamate condensers were supplied
by OLMI (Italy) - not among the approved short listed
vendors
Start-up personnel of TEC did not have previous
experience with ACES process
As a turn-key contractor TEC had the overall
responsibility during execution
Safety
HAZOP of the process was not done by TEC or
BCIC
Safety drills and procedures were not followed
Poor design of control room and inadequate exit
facilities
No arrangement of emergency lights
Aftermath
TEC made the compensation
Modifications were done as recommended to
ensure safety
However, UFFL became a financially lost concern
after the accident
