Biezma - 2020 - Most Fatal Oil & Gas Pipeline Accidents Through History - A Lessons Learned Approach
Biezma - 2020 - Most Fatal Oil & Gas Pipeline Accidents Through History - A Lessons Learned Approach
Most fatal oil & gas pipeline accidents through history: A lessons
T
learned approach
⁎
M.V. Biezmaa, M.A. Andrésb, D. Agudoa, , E. Brizc
a
Materials Science and Engineering Department, University of Cantabria, 39004 Santander, Spain
b
Navigation and Naval Construction Sciences and Techniques Department, University of Cantabria, 39004 Santander, Spain
c
Mechanical Engineering Department, Basque Country University, 48013 Bilbao, Spain
A R T IC LE I N F O ABS TRA CT
Keywords: Pipelines are used throughout the world for oil & gas transportation purposes. This method of
Pipeline transport is reasonably safe, yet accidents keep occurring nowadays. It is important, for the safety
Accident of the oil & gas pipeline network, to remember the history that has led the industry to the actual
Lessons learned approach development point. This research compiles the most fatal oil and gas pipeline accidents through
Risk management
history. Of the compiled accidents, the 10 most fatal are selected for a brief but precise review of
Safety
their root causes and the lessons learned from them. The core objective of this paper is to learn
from the experience of the documented pipeline failures, with the purpose of building a safer and
better future for the oil & gas pipeline transportation network.
1. Introduction
The failure of Oil and Gas Pipelines (OGP) is always an unfortunate incident because of the consequences that it entails: in some
cases, the consequences may be economical, environmental or, in the worst imaginable condition, the accidents can provoke human
losses [1]. The relative high safety that the OGP system has today could not have been achieved without the accidents that happened
through history [2]. The causes that led to these accidents were, in some cases, not known before they took place and, therefore,
could not have been investigated prior to the accident.
OGP can present fatal damage that leads to accidents in the form of a rupture or, more frequently, in the form of latent damage
that can result in failure at a later date [3]. For this reason, the purpose of this paper is to benefit from the fatal OGP failures that have
been documented trough history, in order to avoid the need of further pipeline failures by learning from these unpleasant experi-
ences. Therefore, the approach taken by this paper is considered as a Lessons Learned Approach or LLA [4].
This paper is structured as follows. In Section 2, the failure causes established by the most important pipeline failure databases are
explained, along with the failure criteria adopted for this paper. Subsequently, Section 3 presents a review of the established causes of
pipeline failure and the classifications adopted by the most relevant organizations in the field. Furthermore, the classification as-
sumed by this paper is explained and justified. In Section 4, the pipeline accidents that entailed more than 20 fatalities through
history are compiled, the 10 most fatal OGP accidents are presented, and an analysis of the lessons-learned from them is realized.
Finally, in Section 5, the conclusions of this research can be found.
⁎
Corresponding author at: Escuela Técnica Superior de Náutica, Calle Gamazo, 1, 39004 Santander, Cantabria, Spain.
E-mail address: [email protected] (D. Agudo).
https://doi.org/10.1016/j.engfailanal.2020.104446
Received 20 August 2019; Accepted 11 February 2020
Available online 18 February 2020
1350-6307/ © 2020 Elsevier Ltd. All rights reserved.
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
The construction of the world pipeline infrastructure, initiated with the use of wooden pipes [5]. These primal installations were
first replaced by lead pipes and subsequently by steel pipes in 1920, allowing the use of higher operating pressures [6]. According to
the United States Department of Transportation, more than 50% of the installed pipe system was built between 1950 and 1969, which
implies approximately 228,526,848 km of pipes out of a total of 492,459,264 km of the gas transmission system; about 5% was built
before 1940 and, the rest, from 1970 to the present [7].
In 1960, advances in pipeline technology occurred with the introduction of pressure testing [8], then in the 1970s were im-
plemented the application of thermomechanical processing to substitute the traditional heat treatment techniques in pipeline steels
[9], leading to the development of X70 steel grade pipes [9,10]. According to the API criteria for pipe steel grades classification, the
“X” grade determines the conditions of the pipeline chemical composition and manufacturing, while the numbers following the “X”
establish the minimum yield strength of the pipe in .000 s of psi [11]. Subsequently, accelerated cooling after controlled rolling was
introduced to improve the strength levels (X80 steel) [6,10].
Advances in steel production and rolling practices (hot mechanical treatment, and aimed at the transformation of ingots into
various profiles) occurred in the 1970s and 1980s through the control of the microstructure, further improvements of the rolling
method and chemical additions [12–15]. These resulted in steels with higher yield strengths, which increase mechanical resistance
for equal wall thickness [14,16].
Finally, the transition from the “vintage” to the modern era led to significant changes in the supervision and control of the pipe
system. The first monitoring system was employed in 1965 [17], and, since then, there have been several generations of development.
Sensors, remote controlled valves and specific software can control parameters, such as operating pressure and others, which fa-
cilitate the control of the system and also improve the leak detection and flow control functionalities [16,18–20].
Today, advanced high-strength steels such as APIX80, X100 and X120 are being evaluated for the potential application of high-
pressure and long-distance gas transmission pipelines [10], such as the 3443.99 km pipe network in Alaska whose operating pressure
is 2500 psi [6] or the new Nord Stream 2 1200 km link between Russia and Germany that is to be finished by the end of 2019 [21].
OGP carry the responsibility of transporting hazardous products and, therefore, a failure in an OGP system can lead to significant
consequences such as fire, explosions, environmental damage, and injuries or, in the worst imaginable condition, human casualties
[22].
Various databases that collect data regarding pipeline failure exist throughout the oil & gas industry all over the world, as can be
seen in Fig. 1. In North America, the spearhead OGP incident database is managed by the Pipeline and Hazardous Materials Safety
Administration (PHMSA) [23], which is a branch of the United States Department of Transportation (DOT). In this region, pipeline
operators are obliged by law to report to the DOT every event that involves an undesired release to the environment, while meeting
any of the following criteria [24]:
2
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
In Europe, the reporting of pipeline accidents is not mandatory by law. However, several pipeline operators decided to participate
in the creation of a pipeline incident database in this region, in order to gather data and learn from the incidents that take place. As a
result of this initiative, EGIG (European Gas pipeline Incident data Group) was born [24,25].
There are more OGP failure databases throughout the world, such as UKOPA (United Kingdom Onshore Pipeline Operators’
Association) in the UK, or PID (Pipeline Incident Database) in Canada [24]. In Fig. 1, the official databases that exist nowadays with
the purpose of OGP failure reporting are schematized.
The completely greyed out countries in the world map depicted in Fig. 1 do not have an established database system for reporting
OGP incidents. It is very important to notice that a vast amount of the globe is in this condition, which means that it is easy for OGP
failures that occur to go unnoticed. This is a very unfortunate situation that should be addressed, since learning from OGP failures, as
this paper defends, can prevent future failures that may lead to catastrophic consequences.
The scientific community is not unaware of the importance of having an efficient OGP failure reporting system. For instance,
Castellanos et al. [26] took the OGP failure reporting system a step further, and made the effort of developing a Failure Analysis
Expert System (FAES) database supported by an algorithm through Artificial Neural Network (ANN). The fact that a unified database
is yet to be established hinders the OGP industry need of joining efforts to establish unified criteria that allows the identification of
the variables that induce failure in order to prevent it.
OGP failure is a relatively frequent occurrence [27], and the accidents in this field are usually reported and thoroughly studied,
yet no directives or regulations that determine in a unified manner the way of classifying the failures in OGP exist. One of the most
simplified approaches, as described by Miao and Zhao [28] is to divide the main failure causes of pipeline into four groups: third-
party damage, corrosion, design and construction error, and incorrect operation conditions. These aforementioned authors, report in
their study that corrosion damage and third-party damage contribute the greatest to increment the likelihood of OGP failure.
In the following subsections, the categorization of incident causes that the PHMSA and the EGIG databases adopt aforementioned
databases use are explained, since the criteria is not unanimous. The aforementioned databases have been selected for the criteria
analysis since the authors strongly believe that, nowadays, these are the most impactful incident reporting databases in the OGP
industry.
The PHMSA establishes 8 different categories of pipeline failure causes [29] which are briefly described below:
• Corrosion: deterioration of a metallic pipeline that results from an electrochemical reaction with its environment, resulting in
metal loss in the pipe. The PHMSA differentiates the following specific types of corrosion (the specific types of corrosion not
mentioned below are categorized as “corrosion”):
o External corrosion: corrosion that occurs due to the environmental conditions on the outside of the pipe, as a result of the
interaction of the pipe with the soil, air or water that surrounds it. There are new approaches and techniques to detect this
phenomenon at early stages by means of probabilistic approaches [30–33].
o Internal corrosion: corrosion that occurs due to the environmental conditions inside of the pipeline, as a result of the interaction
of hydrogen sulfide, carbon dioxide, other chemicals, or even water, with the interior walls of the pipe [34–38].
o Stress Corrosion Cracking (SCC): formation of cracks as a result of the interaction of the mechanically stressed pipeline due its
pressurized contents with its environment, that result in a marked decrease of the pipe load bearing capabilities, as have been
very well documented, mainly caused by a mechanism of plasticity-induced crack growth or ductile tearing [39–42].
o Selective Seam Corrosion (SSC): localized corrosion attack along the weld line of pipelines joined by electric resistance welding
(ERW) or electric flash welding (EFW) processes, causing a crevice-like corrosion centered on the bondline of the pipe’s weld
[43,44]. This problem mainly affects pipelines manufactured from 1920 to 1970, since after that, the mentioned pipe joining
techniques were no longer used [45]. It has been reported that SSC has a synergistic effect with fatigue corrosion events [46].
• Excavation damage: failure caused by digging inadvertently into a buried pipeline that sooner or later may lead to the sudden an
unexpected failure of the pipe [47–49].
• Natural force damage: incidents caused by acts of nature such as flooding [50,51], earthquakes [52–54] or lightning [55–57].
• Other outside force damage: incidents caused by vehicle accidents, vandalism, sabotage, terrorism or as a result of another nearby
accident [58–60].
• Material/Weld failure: the pipeline fails due to defects, impurities on the metal, chemical composition and/or improper welding
techniques [61,62].
• Equipment failure: failure of a component or device other than the pipe [63,64].
• Incorrect operation: failure induced by human factors, usually by a mistake made by one of the operators. This category includes
actions like overfilling a tank, leaving a wrong valve opened, overpressuring a piece of equipment or incorrectly marking an area
to be excavated [65,66].
• All other causes: sub-causes out of category.
In Fig. 2, the incident distribution per cause from 2009 to 2018 of the incidents reported to the PHMSA database is shown [67].
3
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 2. PHMSA OGP incident distribution per cause in the 2009–2018 period [67].
Fig. 2 shows that in the USA the most frequent root cause in OGP failures in the last 10 years has been material/weld/equip failure
(43%) by a significant margin, followed by corrosion (17%).
The EGIG database has a smaller pipeline incident cause classification, with 5 defined categories, which are briefly explained
below.
• Corrosion: comprises all of the failures caused by all types of corrosion, although EGIG states that additional information re-
garding the specific corrosion mechanism needs to be provided in the report of each incident.
• External interference: this category includes the failures provoked by digging, piling and ground works. It is very similar to the
“excavation damage” category established by PHMSA.
• Construction defect/material failure: this failure cause is equivalent to the “Material/Weld failure” used by PHMSA.
• Ground movement: includes failures caused by dike break, erosion, flood, landslide [68–70], mining, rivers or unknown but
related to a movement of the terrain in which the pipe is installed.
• Other and unknown: sub-causes out of category such as design error [1], maintenance error, lightning or unknown causes.
Fig. 3 shows the incident distribution per cause from 2007 to 2016 of the incidents reported to the EGIG [27].
As can be seen in Fig. 3, in the countries that participate in the EGIG database external interference has been the most frequent
root cause of OGP pipeline failure (28%) in the 2007–2016 period, closely followed by corrosion (25%). Although it is not specified
amongst the failure causes identified by the EGIG, there is a 4% of the reported failures that were caused by hot tap made by error,
meaning that modifications in the pipeline were made whilst the pipe continued being in operative conditions. There is a notorious
difference between the incident distribution per cause of the incidents in the USA (Fig. 2) and in Europe (Fig. 3), being the only
significant common factor that corrosion is the second most frequent cause in both cases.
Since this study compiles the most lethal pipeline accidents through the world, the authors consider that, with clarification
purposes, utilizing a nomenclature system that adapts to the failures reviewed is the best approach. With regards to the failure causes
criteria adopted for this review, the authors have opted for an intermediate point between the two previously reviewed classification
systems. The reason behind this is that, given the characteristics of the accident data recompiled, the PHMSA approach would be too
detailed, while the EGIG one would not be enough to cover all the particularities. The classification of failure causes adopted for this
review is explained below.
• Vandalism: accidents caused by thieves drilling into the pipe in order to steal fuel.
• Uniform corrosion: failure caused by uniform thinning of the interior or exterior of the pipe due to uniform corrosion.
• Pitting corrosion: failure provoked by pitting corrosion of the pipe.
4
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 3. EGIG OGP incident distribution per cause in the 2007–2016 period [27].
• Third party: failure caused by construction works (digging, piling, ground works).
• Flawed design: pipe failure due to a design mistake.
• Mismanagement: failure induced by human factors, either a mistake by an operator or due to poor decision-making by the
pipeline company.
• Component failure: failure of one of the components of the pipe (e.g. valve, weld). The causes behind the failure are not specified
and cannot be linked to either of the mechanisms exposed before.
• Not defined (ND): there is no available information regarding the causes behind the accident.
4. OGP accidents compilation and analysis
OGP accidents that have been compiled by this research are presented in Table 1, including data such as the date of the accident,
location, number of fatalities (criteria of order), transported fluids, the root cause of the event and the sources from where the data
Table 1
Most fatal OGP accidents through history.
Date Location Fatalities Root cause Substance Sources
5
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 4. Number of fatalities caused by the most lethal OGP accidents ordered chronologically.
was obtained. The criteria adopted by the authors was to take the highest published number, considering different number of
reported fatalities depending on the source: in this paper, 23 accidents were compiled, which caused a total of 4329 fatalities.
As can be seen in Table 1, Nigeria is very present in the top 10 of the most lethal OGP accidents through history, accounting for
the 43% of the total number of fatalities. Trying to remedy these kind of situations, great effort has been put recently towards
mitigating the frequency of OGP failures in Nigeria [71].
In Fig. 4, the number fatalities of caused by the compiled OGP accidents is chronologically ordered by year.
From Table 1, it is indicated that the major proportion of the fatalities is located in the period 1984–2006, where almost 80% of
Fig. 5. Distribution of the fatalities of the most lethal OGP accidents by root cause.
6
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Table 2
Lessons learned from the 10 most fatal OGP accidents through history.
Accident (Fatalities) Lessons learned
Lagos, 1998 (1078) • Pipeline incidents must always be notified immediately to the emergency services and the affected population, in order to reduce
their impact and the number of fatalities.
• Lack of maintenance in significantly poor areas, can induce failure that leads to scavenging, which increments the impacts of
pipeline breakdown. For this reason, properly designed and applied maintenance plans are always of vital paramount.
San Juanico, 1984 (650) • The use of state-of-the-art effective leak detection systems are crucial for the safety of gas and oil transport via pipelines. If the
leak is quickly detected, then the probabilities of escalation are significantly diminished.
• Material selection during design must take into consideration the potential of the pipe being subjected to suffer pitting corrosion.
Therefore, a material that is not prone to pitting corrosion should be prioritized during the design stages in order to avoid accidents
like this.
• Proper isolation techniques should be always designed with the purpose of containing possible accidents. This was not the case in
this accident, allowing the initial blast propagation and causing a tremendous number of fatalities.
Ufa, 1989 (643) • Maintaining an intensive train schedule was prioritized over verifying the information given by a train driver pointing out that he
noticed a very strong gas odor when transiting through the accident area. Therefore, it is sensible to state that human safety should
be always situated over any kind of profit, since the consequences, as happened in this case, can be catastrophic.
• Proper real-time telemetry information is vital to prevent confusions as the one that increased the impact of this accident. Not
having enough data lead the operator to not be able to make an informed decision before raising the pressure of the pipe.
Lagos, 2006 (466) • Same as in the Lagos, 1998 accident.
Texas, 1937 (309) • This accident caused one of the most important improvements towards ensuring gas transportation and consumption safety, due
to the fact that, as result of this disaster, the mixing of “malodorants” with natural gas was made mandatory in the state of Texas,
with other states quickly following [82].
• Although several gases were added “malodorants” as a consequence of this accident, there are still some that are completely
unscented such as carbon monoxide, which in high enough concentrations, is highly lethal. For this reason, installing gas detectors
in public buildings, such as schools, and even in private homes, is advisable.
Guadalajara, 1992 (252) • When flammable materials such as gas & oil are involved, the risk of explosion should always be properly assessed before stating
that no measures are to be taken. Moreover, when the risk of human casualties is involved precautions should be extreme.
• Design is one of the most important stages toward accident prevention since, as happened in this case, a design flaw can cause
severe consequences. For this reason, it is very important that designers have all the required information and know-how in order
to be able to generate the most appropriate design.
Lagos, 2006 (143) • Same as in the accident occurred in Lagos, 1998.
Abia, 2003 (105) • The response made by the company responsible of the pipeline when an incident is reported must be as quick and effective as
possible, and the potential risks can never be disregarded.
• Same as in the accident occurred in Lagos, 1998.
Lagos, 2008 (100) • Construction work should always be properly planned. Before starting any kind of digging, the responsible construction company
must always gather all the updated information regarding the piping systems that run through the area.
Nairobi, 2011 (100) • Exactly determining the failure mechanisms that led to any accident or incident is vital in order to learn from them. If this is not
done, the opportunity of reducing the likelihood of a similar accident happening again is discarded, placing human lives at risk.
On October 17th, 1998, the most devastating accident on the history of oil and gas pipeline transportation occurred, with
unprecedented mortality and extent of damage [114]. This accident is considered the third most fatal within the energy industries
[72]. The death toll was furthermore increased due to the fact that farmers realized that the pipeline was leaking and notified other
nearby villagers, who traveled to the pierced pipe to scavenge for fuel [72]. This fact was proven because many of the dead were
holding a bucket, verifying that theft was going on during the incident, although some sources signal that it might have been possible
that the failure was induced due to lack of maintenance and neglect [84]. Moreover, the government and oil companies involved did
not make public the incident upon realization, causing a delay on the medical assistance arrival, which only happened after many had
already died [72]. In Fig. 6, a picture of the aftermath of this accident is shown, where the buckets that were being used by the
scavengers can be seen in the foreground of the image [74].
The most severe LPG (liquefied petroleum gas) accident through history happened when a 12-inch pipeline transporting LPG
7
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 6. Firefighters trying to extinguish a fire after the 1998 pipeline blaze in Nigeria [74].
ruptured due to pitting corrosion, although the available information states that the causes of the accident are yet to be precisely
defined. The initial blast generated a series of chain-reaction explosions, which, combined with the fact that the pipeline crossed
through a residential complex in the area of the failure, caused an impacting death toll [75]. Fig. 7 depicts the tanks blown in the San
Juanico storage plant during the chain of explosions [77].
The third most impacting OGP accident regarding the number of casualties occurred during the night of the 4th of June 1989,
when two passenger trains were crossing through a zone of natural gas contamination caused by a leak on the Western Siberia/Ural/
Volga pipeline. A spark from one of the train motors caused an explosion equivalent to 300 tons of TNT, which combined with the fact
that the trains affected by the accident were carrying a combined figure of over 1.300 passengers, caused the number of fatalities to
skyrocket. In the literature, mismanagement is signaled as the primary cause of the incident, because the investigation following the
accident concluded that, when the operators of the pipeline realized that the pressure through the line was dropping, decided to raise
said pressure in order to compensate and maintain system stability, causing the broken section of the pipe to increase. This decision
was made mainly because they lacked the resources and telemetry systems to detect a leak and there were no signals of the pressure
drop being dangerous [115]. Fig. 8 depicts one of the trains involved in the accident in the derailed condition after the explosion
[78].
This is the second OGP accident amongst the 4 more lethal that is located in the area of Lagos, Nigeria. The information regarding
this accident is very limited, although the sources signal that the cause of the accident was very likely the same as in the most lethal
pipeline accident through history: pipeline was vandalized with oil-theft purposes, leading more scavengers to turn up to the van-
dalized area and increasing the number of fatalities [84]. The remnant fire after the OGP accident can be seen in Fig. 9 [81].
Fig. 7. Tanks blown in the San Juanico LPG storage plant after the chain of explosions [77].
8
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 8. One of the trains derailed after the Ufa pipeline disaster [78].
Fig. 9. On the background, smoke and flames after the December 6th pipeline accident [81].
It was a regular afternoon in the Consolidated High School, located in New London, Texas, when, several minutes before the end
of the school day, the building was completely destroyed due to an explosion provoked by an undetected natural gas leak, causing the
death of 309 people, of which 294 were students [72]. The investigation about the accident concluded that the cause of the explosion
was that an electric wood-shop sander sparked unscented gas that had accumulated beneath and inside the walls of the school [82].
Fig. 10 depicts the aftermath of the New London School explosion, where the demolished building can be observed in the background
of the image [82].
Once again, an initial explosion causing a chain-reaction of several explosion was the recipe for a high death toll accident. The fact
that impacts the most about this accident is that the gas leakage was noticed by the neighbors, who notified the company and led to a
city worker being sent to inspect the area, discovering high amounts of gasoline fumes. However, the city major deemed that there
was no risk of explosion and the evacuation was discarded [84]. The installation of a zinc-coated pipe with sewer purposes too close
to the existing gasoline pipeline created the environment necessary for galvanic corrosion to occur, which led to the thinning of the
gasoline pipe due to corrosion and the subsequent leak that led to the accident. The large number of casualties was caused because
the highly combustible gas dispersed over a highly populated urban area [98]. The tremendous damage caused by the chain of
9
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
This accident was determined to have been caused by the pressurized petrol pipeline being previously punctured by thieves [84],
and, therefore, vandalism is again the main reason of a lethal OGP accident located in Nigeria. The diesel that was leaking from the
vandalized pipeline caught fire, destroying three nearby villages [72].
This accident, as the ones reviewed in Sections 4.1, 4.4 and 4.7, was caused by vandalism with thieving purposes. The case,
however, has a deeper background, since the puncturing of the pipeline was astonishingly done two months prior to the accident,
without the responsible company doing anything to remediate it, so the accident was far from instantaneous. Furthermore, locals
claimed that they had informed the owner of the leak several days before the accident, but they failed in detaining the leak [84].
The first accident on the list caused by third party occurred when a pipeline transporting diesel exploded in Ijegun, Lagos on May
2006, after it was damaged during construction works when an operator struck the pipeline with a bulldozer, causing it to rupture
and subsequently explode. The explosion damaged the local Primary School and over 15 homes [84]. Fig. 12 depicts the firefighters
trying to extinguish the fire following the OGP pipeline explosion [89].
It was September 12th, 2011, when a pipeline operated by the Kenya Pipeline Company (KPC), which had been leaking oil
Fig. 11. Damage caused by the Guadalajara 1992 pipeline explosions [85].
10
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
Fig. 12. Firefighters trying to put down the fire after the OGP pipeline explosion [89].
previously, caught fire. The high death toll was caused, once again, due to the fact that locals were visiting the area of the incident to
collect the leaking oil. A valve that had failed under pressure was signaled as the location of the leak [84]. However, the causes that
led the valve to fail are not explained. In this kind of accidents, it is of paramount importance to pinpoint the reasons behind the valve
failure. The reasoning behind this statement, is that it is highly likely that more valves than the one that failed are installed through
the pipeline system, either in the same pipeline that failed, or in another transport systems throughout the world, so the failure could
potentially reproduce. It is very different that the failure was induced by a design flaw on the valve, by some corrosion mechanism or
due to subjecting the valve to a pressure higher than its capabilities. The gruesome consequences of this accident can be seen in Fig. 1,
where a man is sitting devastated in the foreground of the picture after discovering that two of his children have perished as a result
of the OGP failure [90] (see Fig. 13).
Table 2 presents, adopting a Lessons Learned Approach, LLA, by the authors, all the lessons learned from the most lethal OGP
accidents through history that this paper has compiled and analyzed in the previous subsections.
5. Conclusions
OGP fatal accidents as the ones reviewed in this paper are the most undesirable and unfortunate incidents that can occur within
the industry. However, the lessons learned from them play a critical role on the history and development of the pipeline transport
system.
This paper has compiled a total of 23 OGP fatal accidents with more than 20 deaths that have occurred trough history, with a
combined number of 4329 fatalities. Vandalism has proven to be the most life-detrimental root cause in these accidents, being the
responsible for 43% of the fatalities, while mismanagement and pitting corrosion are neck-and-neck for second place, accounting
each for 15% of the total number of deaths. The fact that the root cause behind the accidents responsible of 9% of the fatalities has not
Fig. 13. A man sits in a state of shock after discovering the remains of two of his children after the 2011 Kenya pipeline fire [90].
11
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
yet been defined is a notable and preoccupying observation. Moreover, Nigeria has a prominent presence in the top 10 of the most
lethal OGP accidents through history: 43% of the total number of fatalities occurred in this country.
The LLA taken to analyze the 10 most lethal OGP accidents has proportioned more than 15 lessons learned that are stated in this
paper and pursue the objective of being always present for the OGP manufacturers, operators, and for the general population. The
general population has had a significate influence in the most lethal OGP accidents and, therefore, should be aware of the risks that
OGP transportation implies.
As a final conclusion, the authors state that it is highly advisable to benefit from the experience of the documented OGP pipeline
failures, learning from the past, with the purpose of building a safer and better future.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
References
[1] J.G. Ramírez-Camacho, et al., Assessing the consequences of pipeline accidents to support land-use planning, Saf. Sci. 97 (2017) 34–42, https://doi.org/10.
1016/j.ssci.2016.01.021.
[2] G. Gabetta, G. Gori, The use of knowledge management to improve pipeline safety, Integrity of Pipelines Transporting Hydrocarbons, Springer, Dordrecht,
2011, pp. 1–16, , https://doi.org/10.1007/978-94-007-0588-3_1.
[3] W. Worthington, Monitoring for transient pressures in pipelines, Pipelines 2005: Optimizing Pipeline Design, Operations, and Maintenance in Today's Economy,
2005, pp. 886–898, , https://doi.org/10.1061/40800(180)71.
[4] N. Milton, The Lessons Learned Handbook: Practical Approaches to Learning from Experience, Elsevier, 2010.
[5] J.E. Brantly, History of Oil Well Drilling, Gulf Publishing Company, Book Division, 1971.
[6] M. Torres, Microstructural Investigation of Vintage Pipeline Steels Highly Susceptible to Stress Corrosion Cracking, The University of Texas at El Paso, 2016.
[7] J.F. Kiefner, M.J. Rosenfeld, The role of pipeline age in pipeline safety. INGAA, Oct, 2012. Available from: https://www.ingaa.org/file.aspx?id=19307.
[8] H.A. Kishawy, H.A. Gabbar, Review of pipeline integrity management practices, Int. J. Press. Vessels Pip. 87 (7) (2010) 373–380, https://doi.org/10.1016/j.
ijpvp.2010.04.003.
[9] I.De S. Bott, et al., High-strength steel development for pipelines: a Brazilian perspective, Metall. Mater. Trans. A 36 (2) (2005) 443–454, https://doi.org/10.
1007/s11661-005-0315-9.
[10] D.B. Rosado, et al., Latest developments in mechanical properties and metallurgical features of high strength line pipe steels, Int. J. Sustain. Constr. Des. 4 (1)
(2013), https://doi.org/10.21825/scad.v4i1.742.
[11] API (American Petroleum Institute), API Specification 5L, 46th ed. Washington DC, 2018.
[12] S.K. Biswas, S. Chen, A. Satyanarayana, Optimal temperature tracking for accelerated cooling processes in hot rolling of steel, Dyn. Control 7 (4) (1997)
327–340, https://doi.org/10.1023/A:1008268310234.
[13] A.J. Deardo, Accelerated cooling: a physical metallurgy perspective, Proceedings of the Metallurgical Society of the Canadian Institute of Mining and
Metallurgy, Pergamon, 1988, pp. 3–27, , https://doi.org/10.1016/b978-0-08-035770-6.50006-4.
[14] M. Pontremoli, et al., Development of grade API X80 pipeline steel plates produced by controlled rolling, Met. Technol. 11 (1) (1984) 504–514, https://doi.org/
10.1179/030716984803275197.
[15] M.V. Biezma, J.R. San Cristóbal, Letter to the editor: Is the cost of corrosion really quantifiable? Corrosion 62 (12) (2006) 1051–1055, https://doi.org/10.5006/
1.3278238.
[16] R.W. Revie (Ed.), Oil and Gas Pipelines: Integrity and Safety Handbook, John Wiley & Sons, 2015.
[17] J.F. Kiefner, C.J. Trench, Oil Pipeline Characteristics and Risk Factors: Illustrations from the Decade of Construction, American Petroleum Institute, 2001.
[18] J. Aleksandersen, et al., The smart plug: a remotely controlled pipeline isolation system, The Eleventh International Offshore and Polar Engineering Conference,
International Society of Offshore and Polar Engineers, 2001.
[19] M.A. Westhoff, Using operating data at natural gas pipelines, Proceedings: International Symposium on Transportation Recorders, (1999).
[20] I. Stoianov, et al., PIPENETa wireless sensor network for pipeline monitoring, Proceedings of the 6th International Conference on Information Processing in
Sensor Networks, 2007, pp. 264–273, , https://doi.org/10.1145/1236360.1236396.
[21] A. Goldthau, Assessing Nord Stream 2: regulation, geopolitics & energy security in the EU, Central Eastern Europe & the UK, European Center for Energy and
Resource Security. Strategy Paper vol. 10, (2016).
[22] W.K. Muhlbauer, Pipeline Risk Management Manual, Gulf Publishing Company, 1996.
[23] Pipeline and Hazardous Materials Safety Administration. PHMSA [online], 2019. Available from: https://www.phmsa.dot.gov/ (accessed 11 March 2019).
[24] R. Bolt, Using or creating incident databases for natural gas transmission pipelines. Report of study group 3.4, 23rd World Gas Conference, Amsterdam, The
Netherlands, 2006, https://doi.org/10.1115/ipc2006-10619.
[25] European Gas pipeline Incident data Group, EGIG [online], 2019. Available from: https://www.egig.eu (accessed 15 March 2019).
[26] V. Castellanos, et al., Failure analysis expert system for onshore pipelines. Part–I: Structured database and knowledge acquisition, Expert Syst. Appl. 38 (9)
(2011) 11085–11090, https://doi.org/10.1016/j.eswa.2011.02.153.
[27] European Gas pipeline Incident data Group, 10th Report of the European Gas pipeline Incident data Group [online], 2018. Available from: https://www.egig.
eu/reports (accessed 20 June 2019).
[28] C. Miao, J. Zhao, Risk analysis for the urban buried gas pipeline with fuzzy comprehensive assessment method, J. Press. Vessel Technol. 134 (2) (2012), https://
doi.org/10.1115/1.4004625.
[29] Pipeline Failure Causes, PHMSA [online], 2019. Available from: https://www.phmsa.dot.gov/incident-reporting/accident-investigation-division/pipeline-
failure-causes (accessed 11 March 2019).
[30] M.V. Biezma, D. Agudo, G. Barron, A fuzzy logic method: predicting pipeline external corrosion rate, Int. J. Press. Vessels Pip. 163 (2018) 55–62, https://doi.
org/10.1016/j.ijpvp.2018.05.001.
[31] I. Bertuccio, M.V. Biezma, Risk assessment of corrosion in oil and gas pipelines using fuzzy logic, Corros. Eng., Sci. Technol. 47 (7) (2012) 553–558, https://doi.
org/10.1179/1743278212y.0000000028.
[32] B.T. Bastian, et al., Visual inspection and characterization of external corrosion in pipelines using deep neural network, NDT E Int. (2019) 102–134, https://doi.
org/10.1016/j.ndteint.2019.102134.
12
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
[33] H.P. Hong, Inspection and maintenance planning of pipeline under external corrosion considering generation of new defects, Struct. Saf. 21 (3) (1999)
203–222, https://doi.org/10.1016/s0167-4730(99)00016-8.
[34] E.O. Obanijesu, et al., Hydrate formation and its influence on natural gas pipeline internal corrosion rate, SPE Oil and Gas India Conference and Exhibition,
Society of Petroleum Engineers, 2010, , https://doi.org/10.2118/128544-ms.
[35] E.O. Obanijesu, Modeling the H2S contribution to internal corrosion rate of natural gas pipeline, Energy Sour. Part A 31 (4) (2009) 348–363, https://doi.org/
10.1080/15567030701528408.
[36] A.M. Halvorsen, et al., pH stabilization for internal corrosion protection of pipeline carrying wet gas with CO2 and acetic acid, CORROSION 2003, NACE
International, 2003.
[37] M.N. Ilman, et al., Analysis of internal corrosion in subsea oil pipeline, Eng. Fail. Anal. 2 (1) (2014) 1–8, https://doi.org/10.1016/j.csefa.2013.12.003.
[38] C.R. Azevedo, Failure analysis of a crude oil pipeline, Eng. Fail. Anal. 14 (6) (2007) 978–994, https://doi.org/10.1016/j.engfailanal.2006.12.001.
[39] W. Zheng, Effects of hydrostatic testing on the growth of stress-corrosion cracks, 1998 2nd International Pipeline Conference, American Society of Mechanical
Engineers, 1998, pp. 459–472, , https://doi.org/10.1115/IPC1998-2053.
[40] J. Li, et al., Investigation of plastic zones near SCC tips in a pipeline after hydrostatic testing, Mater. Sci. Eng., A 486 (1–2) (2008) 496–502, https://doi.org/10.
1016/j.msea.2007.09.046.
[41] S.Sh. Abedi, A. Abdolmaleki, N. Adibi, Failure analysis of SCC and SRB induced cracking of a transmission oil products pipeline, Eng. Fail. Anal. 14 (1) (2007)
250–261, https://doi.org/10.1016/j.engfailanal.2005.07.024.
[42] J. Luo, et al., Stress corrosion cracking behavior of X90 pipeline steel and its weld joint at different applied potentials in near-neutral solutions, Nat. Gas Ind. B 6
(2) (2019) 138–144, https://doi.org/10.1016/j.ngib.2018.08.002.
[43] S.J. Luo, R. Wang, Identification of the selective corrosion existing at the seam weld of electric resistance-welded pipes, Corros. Sci. 87 (2014) 517–520, https://
doi.org/10.1016/j.corsci.2014.06.044.
[44] G.T. Quickel, B.C. Rollins, J.A. Beavers, Analysis of seam weld related pipeline failures, Mater. Sci. Technol. 8 (2008) 514–523.
[45] J.F. Kiefner, Dealing with low-frequency-welded ERW pipe and flash-welded pipe with respect to HCA-related integrity assessments, ASME 2002 Engineering
Technology Conference on Energy, American Society of Mechanical Engineers, 2002, pp. 699–706, , https://doi.org/10.1115/etce2002/pipe-29029.
[46] M.D. Chapetti, J.L. Otegui, J. Motylicki, Fatigue assessment of an electrical resistance welded oil pipeline, Int. J. Fatigue 24 (1) (2002) 21–28, https://doi.org/
10.1016/s0142-1123(01)00111-6.
[47] K.A. Macdonald, et al., Assessing mechanical damage in offshore pipelines–Two case studies, Eng. Fail. Anal. 14 (8) (2007) 1667–1679, https://doi.org/10.
1016/j.engfailanal.2006.11.074.
[48] S. Duan, P. Shen, Analysis of nearby pipeline damage induced by deep excavation, Eng. Mech. 4 (2005) 79–83.
[49] C.E. Restrepo, J.S. Simonoff, R. Zimmerman, Causes, cost consequences, and risk implications of accidents in US hazardous liquid pipeline infrastructure, Int. J.
Crit. Infrastruct. Prot. 2 (1–2) (2009) 38–50, https://doi.org/10.1016/j.ijcip.2008.09.001.
[50] S. Girgin, E. Krausmann, Historical analysis of US onshore hazardous liquid pipeline accidents triggered by natural hazards, J. Loss Prev. Process Ind. 40 (2016)
578–590, https://doi.org/10.1016/j.jlp.2016.02.008.
[51] E. Krausmann, et al., Industrial accidents triggered by earthquakes, floods and lightning: lessons learned from a database analysis, Nat. Hazards 59 (1) (2011)
285–300, https://doi.org/10.1007/s11069-011-9754-3.
[52] M.J. O'rourke, Xuejie Liu, Response of buried pipelines subject to earthquake effects, 1999. Available from: http://hdl.handle.net/10477/588.
[53] A. Hindy, M. Novak, Earthquake response of underground pipelines, Earthq. Eng. Struct. Dyn. 7 (5) (1979) 451–476, https://doi.org/10.1002/eqe.
4290070506.
[54] S. Yoon, D.H. Lee, H. Jung, Seismic fragility analysis of a buried pipeline structure considering uncertainty of soil parameters, Int. J. Press. Vessels Pip. (2019),
https://doi.org/10.1016/j.ijpvp.2019.103932.
[55] S.B. Cunha, Comparison and analysis of pipeline failure statistics, 2012 9th International Pipeline Conference, American Society of Mechanical Engineers, 2012,
pp. 521–530, , https://doi.org/10.1115/IPC2012-90186.
[56] P. Venturino, et al., Pipeline failures due to lightning, Eng. Fail. Anal. 64 (2016) 1–12, https://doi.org/10.1016/j.engfailanal.2016.02.021.
[57] G.T. Quickel, J.A. Beavers, Pipeline failure results from lightning strike: act of mother nature? J. Fail. Anal. Prev. 11 (3) (2011) 227–232, https://doi.org/10.
1007/s11668-011-9447-y.
[58] G. Luft, Pipeline sabotage is terrorist's weapon of choice, Pipeline & gas journal 232 (2) (2005) 42–44.
[59] L. Kraidi, et al., Analyzing the critical risk factors associated with oil and gas pipeline projects in Iraq, Int. J. Crit. Infrastruct. Prot. 24 (2019) 14–22, https://doi.
org/10.1016/j.ijcip.2018.10.010.
[60] J.R. Dancy, V.A. Dancy, Terrorism and oil & gas pipeline infrastructure: vulnerability and potential liability for cybersecurity attacks, ONE J 2 (2016) 579
https://digitalcommons.law.ou.edu/onej/vol2/iss6/2.
[61] J.L. Otegui, et al., Weld failures in sleeve reinforcements of pipelines, Eng. Fail. Anal. 8 (1) (2001) 57–73, https://doi.org/10.1016/s1350-6307(99)00049-7.
[62] H.I. Shafeek, et al., Automatic inspection of gas pipeline welding defects using an expert vision system, NDT E Int. 37 (4) (2004) 301–307, https://doi.org/10.
1016/j.ndteint.2003.10.004.
[63] A. Aljaroudi, et al., Risk assessment of offshore crude oil pipeline failure, J. Loss Prev. Process Ind. 37 (2015) 101–109, https://doi.org/10.1016/j.jlp.2015.07.
004.
[64] L. Shi, C. Wang, C. Zou, Corrosion failure analysis of L485 natural gas pipeline in CO2 environment, Eng. Fail. Anal. 36 (2014) 372–378, https://doi.org/10.
1016/j.engfailanal.2013.11.009.
[65] C.R. Azevedo, A. Sinatora, Failure analysis of a gas pipeline, Eng. Fail. Anal. 11 (3) (2004) 387–400, https://doi.org/10.1016/j.engfailanal.2003.06.004.
[66] Z.H. Yang, C. Wang, H.U. Shi-Jie, Study on fuzzy comprehensive evaluation system of city gas pipeline based on AHP, J. Saf. Environ. 2 (2013).
[67] PHMSA, Pipeline Incident 20 Year Trends [online). Available from: https://www.phmsa.dot.gov/data-and-statistics/pipeline/pipeline-incident-20-year-trends
(accessed 15 June 2019).
[68] J.Y. Zheng, et al., Failure analysis and safety evaluation of buried pipeline due to deflection of landslide process, Eng. Fail. Anal. 25 (2012) 156–168, https://
doi.org/10.1016/j.engfailanal.2012.05.011.
[69] E.M. Lee, et al., Landslide-related ruptures of the Camisea pipeline system, Peru, Q. J. Eng. Geol. Hydrogeol. 42 (2) (2009) 251–259, https://doi.org/10.1144/
1470-9236/08-061.
[70] R. Bruschi, et al., Failure Modes for Pipelines in Landslide Areas, American Society of Mechanical Engineers, New York, NY (United States), 1995.
[71] A. Ambituuni, E. Ochieng, J.M. Amezaga, Optimizing the integrity of safety critical petroleum assets: A project conceptualization approach, IEEE Trans. Eng.
Manage. 66 (2) (2018) 208–223, https://doi.org/10.1109/tem.2018.2839518.
[72] B.K. Sovacool, The costs of failure: A preliminary assessment of major energy accidents, 1907–2007, Energy Policy 36 (5) (2008) 1802–1820, https://doi.org/
10.1016/j.enpol.2008.01.040.
[73] The Associated Press, Oil Pipeline Blows Up in Nigeria, Killing 250, The New York Times [online). 19 October 1998. Available from: https://www.nytimes.
com/1998/10/19/world/oil-pipeline-blows-up-in-nigeria-killing-250.html (accessed 1 May 2019).
[74] Nigeria oil blast kills scores, BBC News [online], 2003. Available from: http://news.bbc.co.uk/2/hi/africa/3009756.stm (accessed 5 May 2019).
[75] J. Bhandari, et al., Modelling of pitting corrosion in marine and offshore steel structures–A technical review, J. Loss Prev. Process Ind. 37 (2015) 39–62, https://
doi.org/10.1016/j.jlp.2015.06.008.
[76] G. Arturson, The tragedy of San Juanico - the most severe LPG disaster in history, Burns 13 (2) (1987) 87–102, https://doi.org/10.1016/0305-4179(87)
90096-9.
[77] R. Jimenez, A 25 años de las explosiones en San Juanico huele a gas y a olvido, El Universal [online). Available from: https://archivo.eluniversal.com.mx/
graficos/especial/EU_sanjuanico/index.html (accessed 5 May 2019).
[78] J. Nevett, How the Ufa train disaster was overshadowed by Tiananmen Square, BBC News [online], 2019. Available from: https://www.bbc.com/news/world-
13
M.V. Biezma, et al. Engineering Failure Analysis 110 (2020) 104446
14