METALLURGY OF CORROSION, FAILURE ANALYSIS AND CORROSION PREVENTION

1.0 Introduction
1.1 Corrosion theory:
Corrosion is the disintegration of metal through unintentional chemical or electrochemical action,
starting at its surface. The best known case is the rusting of steel.

Figure-1 Electrochemical process of corrosion

Corrosion process (anodic reaction) of the metal dissolving as ions generates some electrons, as
shown in Figure-1, that are consumed by a secondary process (cathodic reaction). The sites
hosting these two processes can be located close to each other on the metal’s surface, or far apart
depending on the circumstances. This simple observation has a major impact in many aspects of
corrosion prevention and control, for designing new corrosion monitoring techniques to avoid the
most insidious or localized forms of corrosion.

Corrosion processes are usually electrochemical in nature, having the essential features of a
battery. When metal atoms are exposed to an environment containing water molecules they can
give up electrons, becoming themselves positively charges ions provided an electrical circuit can
be completed. This effect can be concentrated locally to form a pit or, sometimes, a crack, or it
can extend across a wide area to produce general wastage.

1.2 Factors affecting corrosion

Following are the different factors affecting corrosion and these factors determine the nature and
extent of destruction by corrosion:
1. Material of construction
2. Presence of inclusions or other foreign matter at surface
3. Homogeneity of material structure
4. Nature of corrosive medium
5. Incidental environmental factors (temperature, presence of oxygen, movement velocity of
medium)
6. Presence of stress (residual or applied, steady or cyclic)
 7. Presence of scales (continuous or broken)
8. Deposits on surfaces (porous or semi-porous)
9. Built-in crevices
10. Galvanic effects between dissimilar metals
11. Presence of stray current from external sources

Figure-2 Chart of different corrosion modes in petroleum and chemical industries

1.3 Forms of corrosion

The destruction by corrosion is dependent on various factors and each factor contributes
individually in a different way for initiation and enhancement of corrosion ultimately leading to
failure of the equipment or component of a unit. It is therefore imperative to understand the
effects of each of the contributing factors on corrosion failures so that the failure sources can be
unearthed. The nature of failure and the corrosion mechanism attributed to the failure classifies
the corrosion types or forms.

1.4 Human error as factor of corrosion failures:

A three year study to determine major failure contribution carried out by a failure analysis
company capturing 131 failures had the following listing –
Number of failures Cause of % of total
failures failures
23 Corrosion 18 %
57 Fatigue 44 %
15 Wear 11 %
17 Corrosion fatigue 13 %
19 Overload 15 %
131 Total failures 100 %
Note: In “Fatigue” failures, corrosion was not needed to affect the failure whereas in “Corrosion
fatigue”, there was cyclic loading but not severe enough to cause cracking without corrosion.

According to this study, the attribution of responsibility for corrosion failures investigated was
broken down into the following –
 Contributing factor % of total failures
Lack of proving: new design, material or 36 %
process
Lack of or wrong specifications 16 %
Bad inspection 10 %
Human error (operational) 12 %
Poor planning or coordination 14 %
Others 4%
Unforeseen 8%
Total 100 %
This data set clearly indicates that only 8% of corrosion failures were unforeseeable. In other
words, 92% of the corrosion failures could be preventable.

Figure-3 Mechanical factors controlling corrosion failure

2.0 Metallurgical factors affecting corrosion:

The structure of metals and alloys is important in determining their corrosion characteristics. For
example, sulphide inclusions in pure iron have a marked tendency to react in even mild corrosive
environment. Also, a material heavily deformed by plastic deformation at room temperature (cold
worked, cold rolled, cold drawn etc.) can have a low corrosion resistance. Figure-3 shows some
metallurgical factors that can affect corrosion failure of metals and alloys.

In deformed metals, the grains are deformed and grain structures are completely disrupted. In this
condition, the material is somewhat more reactive in electrochemical environments. Additionally,
the influence of impurities, inclusions, grain boundaries and differences in grain orientation may
also result significantly different electrochemical reactivates in metals and alloys.

2.1 Effect of crystal structure defects and phases of corrosion:

The crystal structure assumed to be perfect in three dimensions may contain variations caused by
crystalline defects. These defects may be vacancies caused by the absence of atoms in the crystal,
impurity of atoms of different sizes, interstitial atoms and large lattice disturbance called
dislocations. Each of the mentioned imperfections can produce highly localized differences in
 electrochemical behavior and corrosion resistance of the material as well. The vacancy, the
impurity atom and interstitial atom are point defects whereas dislocations are line defects
affecting a much greater volume of crystal. In corroding environment (like petroleum and / or
chemical industries) these areas are usually more anodic than the surrounding matrix. The large
numbers of triangular corrosion pits are a result of electrochemical attack due to stress field
around the dislocation. The shape of the pit is related to the orientation of the grain to the
corroded surface. Tendency of a metal to react in electrochemical environments reduces
proportionately if its purity increases. However, pure metals have low mechanical strength and
impurities or imperfections are inherent cause of corrosion in aggressive environments.

Figure-4 Typical microstructure of 0.4% C steel, revealed ferrite and pearlite.

Existence of more than one phase in an alloy usually results in poorer corrosion resistance than
equivalent single phase materials. Figure-4 shows multiphase steel alloys of different carbon
contents. While designing a corrosion resistant alloy, it is important to have a good combination
of mechanical, physical and fabrication qualities which tend to make the material structurally and
economically useful.

2.2 Effect of composition inhomogeneity in castings on their corrosion:

Inhomogeneity in composition within a casting occurs due to non-equilibrium solidification. A
variation in composition will exist from interior to the surface of the grain if grain of the material
is rapidly solidified from liquid because diffusion was not fast enough. This variation in chemical
composition from interior to exterior of grain is called “coring”. Cored structures have widely
different corrosion characteristics and actually have a built-in electrochemical cell. Many
intergranular corrosions result from this type of chemical inhomogeneity. Reheating and holding
the metal (grain) for relatively long time at temperatures just below the solidus line allows rapid
diffusion and assist in homogenization of the alloy.

2.3 Effect of heat treatment on corrosion behavior:

We have already discussed how homogenizing anneal helps to bring chemical homogeneity of an
alloy by diffusion. Deformation by cold working increases dislocations. Impurities or alloying
elements migrate to these dislocation sites and change the electrochemical characteristics of
 crystals. These areas of high dislocations are thus subjected to pitting corrosion. Full annealing of
a cold worked metal produce a more uniform crystal structure, with fewer defects and
dislocations leading to better corrosion resistance. In precipitation hardened alloys, the
precipitated phase may cause a so-called denuded zone adjacent to grain boundaries that is
preferentially attacked by corrosion. Such type of corrosion is called intergranular corrosion.

2.4 Martensitic transformation and corrosion in steels:

Quenching of steel from gamma region (austenitic phase) results in formation of hard and brittle
martensitic phase. Tempering at intermediate temperature to this steel gives best combination of
strength and toughness. However, without surface protection, these steels are easy prey to stress
corrosion cracking. A good example is that of bare high strength steel used in sucker-rod pumps
which suffers SCC after interacting with salt water or hydrogen sulphide solutions.

2.5 Corrosion properties and material selection:

The optimum properties associated with each material selection criteria can seldom all be found
in a single material, especially when the operating conditions become aggressive. Chrome (0.5 to
9%) – Moly (0.5 to 1%) ferritic alloy steels are most commonly used for temperatures up to
650°C. These low alloy steels have inadequate corrosion resistance to many elevated temperature
environments for which more highly alloyed Ni-Cr-Fe compositions are required. For
applications where carbon or low-alloy steels are not suitable, the most common choice of
material is from within the 18Cr-8Ni austenitic group of stainless steel. These alloys and the
18Cr-12Ni stainless steels are favoured for their corrosion resistance in many environments and
their oxidation resistance at temperatures up to 816°C.

3.0 Corrosion control by application of metallurgical principals:

Following metallurgical principals can be applied to reduce or prevent corrosion:
1. Use of high purity material
2. Use of alloys
3. Effective heat treatment
4. Use of surface coatings
5. Knowledge of metallurgical history of used material (for users)
Purity is to be considered in right combination in context to desired mechanical properties
qualifying for a particular service of material. Reducing sulfur content in steels reduces corrosion
attack on steels. Similarly, lead exceeding 0.002% increases intergranular corrosion in zinc die-
cast alloys by precipitating at grain boundaries. 300 series stainless steel has maximum 0.03% C
which reduces sensitization occurring by welding and/or heat treatment.
Alloying addition like addition of chromium to iron in concentration of more than 12% when a
chromium-iron oxide passive layer imparts the renowned corrosion resistant property to stainless
steel. Aluminum in iron forms an intermetallic FeAl 3 making it resistant to pitting corrosion.
Addition of chromium or magnesium changes FeAl3 to a complex Al-Fe-Mn or Al-Fe-Cr
compound whose potential is close to aluminum itself. So, by changing the electrochemical
potential of the second phase in an alloy, the corrosion resistance can be improved.
Heat treatment has already been discussed as beneficial for chemical homogeneity or uniform
crystal structure in metals thereby reducing the sites of electrochemical characteristics. Heat
 treatment of surfaces to increase surface hardness of material or improve the stability of surface
films improves corrosion resistance but is particularly useful in improving fretting and erosion-
corrosion resistance which is commonly applied in shell or sleeve bearings of rotating equipment.
Surface coatings of various types are used as barriers between substrate and the environment to
reduce chances of reactions that occur in metal-environment interface. These include a wide range
of metals, organic or inorganic materials.
Metallurgical history of the material influences all forms of deterioration by corrosion.
Impurities retained during metal extraction, inclusions or imperfections introduced during casting
and forming and structural variations generated by heat treatment alter the corrosion stability of
metals and alloys. For example steels produced in acid open hearth will have higher sulfur level
and resultantly low resistance to corrosion.

4.0 Stainless steel corrosion:

Stainless steel is mainly used in wet environments like oil fields or chemical industries. While
chromium and molybdenum makes the steel resistant to aggressive medium, nickel reduces the
risk of stress corrosion cracking (SCC). Austenitic stainless steels are more resistant to general,
crevice and pitting corrosion. Along with composition, corrosion resistance in stainless steel is
also a function of heat treatment, surface condition and fabrication procedures.

4.1 Pickling and passivation in stainless steel:

Both pickling and passivation are chemical (acid) treatments applied to the surface of stainless
steel to remove contaminants deposited by manufacturing processes like fabrication etc. and assist
formation of a continuous chromium-oxide passive film.

4.2 Stainless steel weld decay or sensitization:

This is a type of intergranular corrosion which occurs in heat-affected zone of welded and cast
stainless steel components due to precipitation of chromium carbides at grain boundaries during
cooling. This process results in depletion of chromium in immediate neighbor of chromium
precipitated zone arising in local loss of corrosion resistance. This phenomenon can be controlled
by carrying out a post-weld heat treatment to restore uniform composition at grain boundaries or
addition of titanium or niobium which, due to higher affinity to carbon, forms carbides and keeps
chromium free to diffuse uniformly.

4.3 Challenges in welding of stainless steel:

It is important to stress relieve stainless steel welds exposed to environments that could cause
stress corrosion cracking (SCC) resulted from high tensile stress due to large variation in thermal
expansion during solidification of molten pool and base metal. To the extent possible, chemically
and electrochemically similar weld filler metals should be used to avoid phase transformations
during welding and provide corrosion stability to welded structures.

4.4 Challenges in heat treatment of stainless steel:

Chromium in stainless steel, particularly the 300 series, remains tied up as carbide formed by
sensitization at grain boundaries during heat treatment and no longer act as a deterrent to
 corrosion. The grain boundaries stay susceptible to intergranular corrosion and are anodic to the
surrounding grains.

Figure-5 Intergranular corrosion in stainless steel

Sensitized stainless steels can deteriorate completely in strongly acidic solutions. Figure-5 shows
schematic illustration of stainless steel intergranular corrosion.

5.0 Stress Corrosion Cracking (SCC) in steels:

A good example of SCC, which is a complex influence of combined effect of tensile stress and a
corrosive environment, is a buried pipeline. Residual stresses developed by cold forming,
welding, heat treatment and machining tend to approach yield strength and prove deadly in
aggressive environments.

Figure-6 SCC in 316 stainless steel chemical process piping system (X300)

Macroscopically, SCC fractures have a brittle appearance. As detection of such fine cracks is
difficult, SCC is the most catastrophic for of corrosion. Figure-6 illustrates intergranular SCC of
a chemical process piping with crack following the grain boundaries. Chloride SCC in austenitic
stainless steel is characterized by multi-branched “lightning bolt” trans-granular pattern. SCC
cracks grow over a range of velocities (10 -3 to 10mm/h) depending on combination of alloy and
environment involved.

5.1 Pipeline Stress Corrosion Cracking:

 Buried pipelines are subjected to environmental abuse, external damage, coating disbondment,
mill defects, soil movements and third party damage. Characteristically uncommon but high-pH
SCC failures in underground pipelines have occurred in a variety of soil. While coal tar, asphalt
and polyethylene coatings are susceptible to SCC, fusion bonded epoxy hasn’t shown such
susceptibility. Cyclic loading and hydrogen ingress are very important and dangerous factors to
germinate the crack and their propagation.

5.2 Control of Stress Corrosion Cracking:

The most effective means of controlling SCC are: 1) material design with right material; 2)
limitation of stress; 3) removal of critical environmental species like hydroxides, chlorides and
oxygen; 4) avoid stagnant areas and crevices.
Material selection is the first line of defense in controlling SCC where service environments and
entire manufacturing processes are to be taken care of.
Control of material stresses during manufacturing can be done by minimizing generation of
residual stresses during welding or forming processes and relieving such stresses if generated.
Selective stress-relieve annealing in large structures need to be done in a controlled way so as to
avoid creating new regions of high stress.
Control of environment is a direct way to control SCC and can be done by removing or
replacing the service environment that promotes either stress or corrosion.

6.0 Corrosion fatigue:

Corrosion fatigue is the result of combined effect of an alternating or cyclic stresses and a
corrosive environment. The fatigue process is believed to cause rupture of the protective film,
upon which corrosion is accelerated. In such conditions, failure can take place at even lower loads
and much earlier than the designed life. Contrary to pure mechanical fatigue, there is no fatigue
limit load in corrosion-assisted fatigue.

Figure-7 Effect of corrosion atmosphere on fatigue failure stress

Fatigue fractures are brittle and the cracks are most often trans-granular, like SCC, but not
branched. The corrosive environment can cause a faster crack growth or in other words, a
corrosive environment can reduce the fatigue strength drastically. Lowering the cyclic stresses or
adopting other corrosion control measures can control corrosion fatigue.

7.0 Hydrogen embrittlement or hydrogen induced cracking:

 Hydrogen produced by corrosion reactions like rusting, cathodic protection, electroplating,
hardening and/or welding (flux) tends to enter the metal in the form of atom rather than be
evolved entirely as gas. The hydrogen ingress reduces ductility, toughness and tensile strength,
causes cracking and catastrophic brittle failures at stresses much below yield stresses. Initiation of
failures by diffused hydrogen is a complex process and can be triggered by the following
mechanisms –
a. Adsorbed hydrogen species recombine to form hydrogen molecules, creating pressure
from within the metal which can increase to alarming levels
b. Formation of brittle hydrides with parent material allows cracks to propagate in a brittle
fashion
c. Hydrogen enhanced de-cohesion (HEDE) where the strength of the atomic bonds of the
parent material are reduced
d. Stationary dislocations begin to move when molecular hydrogen is dissociated and
absorbed into pre-strained material resulting in enhanced localized deformation
In stainless steel, hydrogen diffuses along the grain boundaries and combines with carbon – the
alloying element in steel, to form methane gas which collects in small voids along grain
boundaries building enormous localized pressure enough to initiate a crack.

8.0 Introduction to corrosion failures:

Corrosion failure analysis involves metallurgical investigations of components, equipment,
metals, alloys, coatings, linings, environmental degradation and abuse, misapplication of the
particular metal and mechanical failure. Studies of failure analysis are particularly strong in the
chemical processing, refining, oil & gas, pulp & paper industries, public utilities and
infrastructure assets. Failure mechanisms evaluated usually include:
1. general corrosion
2. localized corrosion
3. intergranular corrosion
4. weld corrosion
5. stress corrosion cracking
6. fatigue & corrosion fatigue
7. fretting & wear
8. erosion
9. overload
10. brittle fracture
11. hydrogen embrittlement
12. hydrogen sulfide cracking
13. microbiological corrosion
14. oxidation, sulfidation & carburization

Preventing or reducing corrosion processes increases the productivity and efficiency of any plant.
The identification of the factors associated with the forms of corrosion can guide failure
investigators. A listing of the most important factors would ensure awareness of the complexity
and multitude of variables involved. Corrosion failure of metallic materials causes billions of
dollars globally every year. The lack of understanding will not only cost a lot of money it will
 also cost human life. Some of the factors and information to be collected before, during and after
a corrosion failure investigation include the following:
1. Analysis of material failures
2. On-site investigations
3. Welding, soldering & brazing analysis
4. Chemical and mechanical testing (ASTM and NACE)
5. Metallographic analysis and investigations
6. Fracture mechanics
7. Engineering calculations
8. Atmospheric testing
9. Accident investigations and recreation of failure
10. Plastic & rubber investigation
11. Micro- and macro-photography
12. Surface analysis
13. Fractography: Fracture mode determination
14. Engineering mechanics analysis
15. Product specification
16. Corrosion testing & investigations (electrochemical and chemical exposure tests)
17. Root-cause failure determination

Failure itself is a human concept. Materials do not fail in and of themselves. They follow the laws
of nature perfectly. A part or equipment will fail only if it is loaded beyond its strength or made to
perform in environment for which it is not designed.

8.1 Failure analysis in corrosion engineering:

The depth of analysis into the roots of failure is the key to accurately unearthing all of the failure
sources. For better understanding of events and mechanisms, a detailed analysis, depending on
depth and complexity of investigation, is divided in to following three categories-
1. Component Failure Analysis (CFA), which looks at the piece of the equipment that
failed and determines the cause and contributing factors.
2. Root Cause Investigation (RCI) is conducted in much greater depth than the CFA and
goes substantially beyond the physical root of a problem to find the human errors and
material flaws involved but doesn't involve management system deficiencies. RCI's are
generally confined to a single operating unit.
3. Root Cause Analyses (RCA) which includes everything the RCI covers plus the minor
human error causes and, more importantly, the management system problems that allow
the human errors and other system weaknesses to exist. An RCA can sometimes extend to
sites other than the one involved in the original problem.
Failure analysis certainly comes with cost but the benefit is that there is a much more complete
recognition of the true origins thereby preventing future failures by eliminating group of
problems influencing the failure.

8.2 Corrosion risk analysis:

 The increasing technical complexity of engineering systems (Figure-8) with introduction of
advanced materials is paralleled by an increasing awareness of the risks, hazards and liabilities
related to the operation of engineering systems.

Figure-8 Graphical representation of interrelation between a defect, a fault and a failure

However, the increasing cost to replace equipment is forcing people and organizations to extend
the useful life of their systems. Mechanical forces, which have normally little effect on the
general corrosion of metals, can act in synergy with operational environments to cause the most
sudden failures by a localized corrosion mechanism. This necessitates adequate training of
personnel to predict damage caused by operating environments.

9.0 Case studies on corrosion in natural gas pipelines:

9.1 Rupture and Fire Near Carlsbad New Mexico August 19, 2000 - EPNG

Accident Synopsis:
5:26 a.m., Saturday, August 19, 2000 – A 30 inch diameter natural gas transmission pipeline
(from Texas and New Mexico to Arizona and California) operated by El Paso Natural Gas
Company (EPNG) ruptured adjacent to the Pecos River near Carlsbad, New Mexico. The released
gas ignited and burned for 55 minutes. Twelve persons who were camping under a concrete-
decked steel bridge that supported the pipeline across the river were killed and their three vehicles
destroyed. Two nearby steel suspension bridges (one bridge supported a water pipeline and a gas
gathering pipeline) for gas pipelines crossing the river were extensively damaged. According to
EPNG, property and other damages or losses totaled $998,296.

Post-accident on-site inspection:

The force of pipe rupture and ignition of escaping gas created a 51foot-wide crater along the pipe.
A 49-foot section of the pipe was ejected from the crater in three pieces measuring approximately
3 feet, 20 feet, and 26 feet in length.

All three ejected pieces showed evidence of internal corrosion damage, but one of the pieces
showed significantly more corrosion damage than the other two. Pits were visible on the inside
 surface of this piece, and various locations showed significant thinning of pipe wall. At one
location, a through-wall perforation was visible. No significant external corrosion damage was
visible in the three pieces or on the two ends of the pipeline remaining in the crater. Pieces were
cut from the ruptured pipeline segments and shipped to the Safety Board's Materials Laboratory
in Washington, D.C., for further evaluation.

The drip between block valve No. 6 and the rupture site, on examination, showed a blackish oily
powdery grainy material. At about 13 feet from the drip opening, this material filled
approximately 70 percent of the cross-sectional area of the drip. No significant material was
observed in the area just underneath and several inches away from the siphon drain at the closed
end of the drip. No significant internal corrosion was observed in the drip.

Probable cause of failure – internal corrosion:

Interconnecting pits were observed on the inside of the pipe in the ruptured area. Typically, these
pits showed the striations and undercutting features that are often associated with microbial
corrosion. A pit profile showed that chloride concentration (evidence of microbial activity) in the
pits increased steadily from top to bottom. All four types of microbes (sulfate reducing, acid-
producing, general aerobic, and anaerobic) were observed in samples collected from two pit areas
in the piece of pipeline where internal corrosion was discovered after the accident about 2,080
feet downstream of the rupture site. The same was corroborated by the damage morphology and
the corrosion product analysis.

Dissolved O2 in an electrolyte could cause pitting and CO 2 being soluble in water, will form
carbonic acid, which is corrosive to carbon steel. H2S when dissolved in water, forms a weak acid
that could corrode carbon steel. These potentially corrosive constituents were present in the gas
that was being transported in the pipeline. Chlorides (enhances pitting corrosion) were observed
in all corrosion product/deposit samples

Chemical analyses showed that the pH (6.7-6.8) of the liquid collected at the Pecos River
compressor station plant inlet separator scrubber was more acidic than the pH (8.2) of the liquid
collected at Keystone compressor station inlet scrubber. Also, the material collected at pig
receivers (pH 6.2-6.3) and the inside material collected from a low spot on pipeline west of the
rupture (pH - 6.4) were more acidic than the material collected near the siphon drain area of the
drip (pH 8.9). The observed low pH in the samples could be a result of dissolved CO 2, and/or H2S
in the water, and/or intrusion of low-pH ground water into the gas supply.

Conclusion:
Water and contaminants such as chlorides, O 2, CO2, and H2S all likely contributed to the observed
corrosion damage. The Safety Board therefore concludes that the corrosion that was found in line
1103 at the rupture site was likely caused by a combination within the pipeline of microbes and
such contaminants as moisture, chlorides, O2, CO2, and H2S.

9.2 Explosion in railroad tank car at Bogalusa, Louisiana in October 23, 1995 - Gaylord
Accident synopsis:
3:55 p.m., Monday, October 23, 1995 – Yellow-brown vapors began leaking from the dome of a
railroad tank car that contained a mixture of nitrogen tetroxide (which is a liquefied poisonous
gas and oxidizer) and water stationed at the Gaylord Chemical Corporation plant in Bogalusa,
Louisiana, USA. The head on the B-end of the tank car failed about 4:45 p.m., resulting in one
end of the tank car jacket being torn away and thrown about 350 feet. The tank car was then
propelled 35 feet down the track and derailed at a track bumping block. Vapors continued to be
released from the opening in the tank car for another 36 hours until the chemical reaction that had
occurred within the tank was brought under control through neutralization and dilution. 3,000
people were evacuated from the area as a result of the vapor cloud. Of 4,710 people who were
treated at local hospitals, 81 people were admitted.

Events preceding the accident:

1. 14th Sept.’95 – Nitrogen tetroxide vapour leaked from a valve on tank car. A mechanic
sprayed water to suppress the vapour and then directed water jet on valve stem
2. 12th Oct.’95 – During transfer of nitrogen tetroxide from car to a storage tank, the
material in storage tank was also being transferred simultaneously.
3. 12th Oct.’95 – Process interlocks shut the chemical reactor down later during the day.
4. 12th & 13th Oct.’95 overnight – Meter indicated that 10100 gallons of product were
transferred from tank car in to two cargo tanks. When transfer stopped, a sample (green
liquid with foam) was collected.
5. 13th Oct.’95, 4am – Water was added to tank car which contained small amount of
residual material. Gaylord technician believed that the small residual content could not be
pushed through eduction pipes.
6. 19th Oct.’95 – Samples from tank car and three cargo tanks were taken for chemical
analysis.
7. 20th Oct.’95 – Gaylord employees began emptying the material from the tank car into the
plant sewer system; after 2 or 3 minutes, a large vapor cloud was released, and the
unloading was stopped.
8. 21st Oct.’95 – A Gaylord employee twice added water to dilute the material in the tank car
and vented the vapors through the scrubber. The internal pressure of the tank car
increased each time the water was added
9. 23rd Oct.’95 – Gauge on the tank car indicated that the internal pressure was 18 psig.
Plans were made to complete the emptying and the cleaning of the tank car.
10. 23rd Oct.’95 12:54pm – Gaylord receives results of the chemical analysis of the sample
collected on 19th Oct. The results indicated that the material was wet nitrogen tetroxide.
These results were unexpected because Gaylord believed that on October 13 the nitrogen
tetroxide had been unloaded from the tank car and that the residue had been diluted with
water. After discussion with laboratory technician, Gaylord personnel concluded that the
sample was not representative.
11. 23rd Oct.’95, 01:30pm – Gaylord employees add more water in to the tank car and the
internal pressure rose from 18 to 80 psig in 5 minutes. Here, the water was turned off. To
vent vapours from tank car, plant scrubber was still being used.
 12. 23rd Oct.’95, 02:30pm – Internal tank car pressure reached 55psig and fire hoses were
reopened when pressure rose to 100 psig, the maximum that the gauge could display.
Water was turned off and vapours again vented out. The pressure remained above
measurement limit of gauge till 3:30pm.
13. 23rd Oct.’95, 03:55pm – A yellow-brown vapour started leaking from tank car.
14. 23rd Oct.’95, 04:45pm – B-end of tank car failed releasing huge reddish-brown vapour
cloud.

Post-accident on-site inspection:

Post-accident examination of the safety relief mechanism mounted in the dome revealed that it
had recently activated. (Activation pressure is about 375 psig.) The examination also revealed an
8-inch-high by 66-inch-wide corroded opening in the B-end head of the tank car near the top. In
addition, at least three distinctive horizontal bands of corrosion were found near the top of the
interior tank wall; one band was at the same level as the opening in the tank head.

Probable cause of failure – non-availability of SOP:

The National Transportation Safety Board (NTSB) determines that the probable cause of the
accident was lack of adequate procedures on the part of the Gaylord Chemical Corporation and
the Vicksburg Chemical Company (testing lab agency) to prevent or detect the contamination of
nitrogen tetroxide with water, resulting in the formation of an extremely corrosive product and the
subsequent failure of the tank car. Contributing to the severity of the accident were the Gaylord
Chemical Corporation's inadequate procedures for emergency transfer of contaminated cargo
from the tank car