International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 2
Number 4
December 2008
Human Errors and Oil Pollution from Tankers
P. Arsenie & R. Hanzu-Pazara
Constanta Maritime University, Constanta, Romania
ABSTRACT: The economical development of the world is based on transportation system. More than half of
the products transported all over the world are carried by sea. Sea transportation is made with different kind of
ships, as bulk carriers, cargo vessels, container ships, tankers. Ships are managed by people. In group or as
individual, anybody can make errors. In maritime area these errors have as results accidents and disasters.
Many of these events affect especially the environment. As 80% of necessary petroleum products are
transported by sea, the risk of a major environment disaster caused by human errors is high. Anyway, over
99% of petroleum cargo transported by sea is carried without incidents. This paper presents the effects of
human errors, mostly cases that involved tankers, which were produced in the navigation and operational
Over the last 40 years or so, the shipping industry
has focused on improving ship structure and the
reliability of ship systems in order to reduce
casualties and increase efficiency and productivity.
We’ve seen improvements in hull design, stability
systems, propulsion systems, and navigational
equipment. Today’s ship systems are technologically
advanced and highly reliable.
Yet, the maritime casualty rate is still high. Why?
Why is it, with all these improvements, we have not
significantly reduced the risk of accidents? It is
because ship structure and system reliability are a
relatively small part of the safety equation. The
maritime system is a people system, and human
errors figure prominently in casualty situations.
About 75-96% of marine casualties are caused, at
least in part, by some form of human error. Studies
have shown that human error contributes to: 84-88%
of tanker accidents, 79% of towing vessel groundings,
89-96% of collisions, 75% of allisions. 75% of fires
and explosions.
Therefore, if we want to make greater strides
towards reducing marine casualties, we must begin
to focus on the types of human errors that cause
A recent study of 100 marine casualties found
that the number of causes per accident ranged from 7
to 58, with a median of 23. Minor things go
wrong or little mistakes are made which, in and
of themselves, may seem innocuous. However,
sometimes when these seemingly minor events
converge, the result is a casualty. In the study,
human error was found to contribute to 96 of the 100
accidents. In 93 of the accidents, multiple human
errors were made, usually by two or more people,
each of whom made about two errors apiece.
But here is the most important point: every human
error that was made was determined to be a
necessary condition for the accident. That means
that if just one of those human errors had not
occurred, the chain of events would have been
broken, and the accident would not have happened.
Therefore, if we can find ways to prevent some of
these human errors, or at least increase the
probability that such errors will be noticed and
corrected, we can achieve greater marine safety and
fewer casualties.
The maritime system is a people system. People
interact with technology, the environment, and
organizational factors. Sometimes the weak link is
with the people themselves; but more often the weak
link is the way that technological, environmental, or
organizational factors influence the way people
perform. Let’s look at each of these factors.
First, the people. In the maritime system this
could include the ship’s crew, pilots, dock workers,
Vessel Traffic Service operators, and others. The
performance of these people will be dependent on
many traits, both innate and learned. As human
beings, we all have certain abilities and limitations.
For example, human beings are great at pattern
discrimination and recognition. There isn’t a machine
in the world that can interpret a radar screen as well
as a trained human being can. On the other hand, we
are fairly limited in our memory capacity and in our
ability to calculate numbers quickly and accurately-
-machines can do a much better job. In addition to
these inborn characteristics, human performance is
also influenced by the knowledge and skills we have
acquired, as well as by internal regulators such as
motivation and alertness.
The design of technology can have a big impact
on how people perform. For example, people come
in certain sizes and have limited strength. So when a
piece of equipment meant to be used outside is
designed with data entry keys that are too small and
too close together to be operated by a gloved hand,
or if a cutoff valve is positioned out of easy reach,
these designs will have a detrimental effect on
performance. Automation is often designed without
much thought to the information that the user needs
to access.
Critical information is sometimes either not
displayed at all or else displayed in a manner which
is not easy to interpret. Such designs can lead to
inadequate comprehension of the state of the system
and to poor decision making.
The environment affects performance, too. By
“environment” we are including not only weather
and other aspects of the physical work environment
(such as lighting, noise, and temperature), but also
the regulatory and economic climates.
The physical work environment directly affects
one’s ability to perform. For example, the human
body performs best in a fairly restricted temperature
range. Performance will be degraded at temperatures
outside that range, and fail altogether in extreme
High sea states and ship vibrations can affect
locomotion and manual dexterity, as well as cause
stress and fatigue. Tight economic conditions can
increase the probability of risk-taking (e.g., making
schedule at all costs).
As you can see, while human errors are all
too often blamed on “inattention” or “mistakes” on
the part of the operator, more often than not they
are symptomatic of deeper and more complicated
problems in the total maritime system. Human
errors are generally caused by technologies and
environments which are incompatible in some way
with optimal human performance.
These incompatible factors “set up” the human
operator to make mistakes. So what is to be done to
solve this problem? Traditionally, management has
tried either to cajole or threaten its personnel into not
making errors, as though proper motivation could
somehow overcome inborn human limitations. In
other words, the human has been expected to adapt
to the system. This does not work. Instead, what
needs to be done is to adapt the system to the human.
The discipline of human factors is devoted to
understanding human capabilities and limitations,
and to applying this information to design
equipment, work environments, procedures, and
policies that are compatible with human abilities. In
this way we can design technology, environments,
and organizations which will work with people to
enhance their performance, instead of working
against people and degrading their performance.
This kind of human-centered approach (that is,
adapting the system to the human) has many benefits,
including increased efficiency and effectiveness,
decreased errors and accidents, decreased training
costs, decreased personnel injuries and lost time, and
increased morale.
What do we mean by “human error”? Human error
is sometimes described as being one of the
following: an incorrect decision, an improperly
performed action, or an improper lack of action
(inaction). Probably a better way to explain human
error and their effects results in environmental
damage, as oil pollution is to provide examples from
two real marine casualties.
The first example is the grounding of the
TORREY CANYON. Again we have clear, calm
weather this time it was a daylight transit of the
English Channel.
While proceeding through the Scilly Islands, the
ship ran aground, spilling 100,000 tons of oil.
At least four different human errors contributed to
this accident.
The first was economic pressure, that is, the
pressure to keep the schedule (pressure exerted on
the master by management).
The TORREY CANYON was loaded with cargo
and headed for its deep-water terminal in Wales.
The shipping agent had contacted the captain to
warn him of decreasing tides at Milford Haven, the
entrance to the terminal. The captain knew that if he
didn’t make the next high tide, he might have to wait
as much as five days before the water depth would
be sufficient for the ship to enter. This pressure to
keep to schedule was exacerbated by a second factor:
the captain’s vanity about his ship’s appearance.
He needed to transfer cargo in order to even out
the ship’s draft. He could have performed the
transfer while underway, but that would have
increased the probability that he might spill a little
oil on the decks and come into port with a “sloppy
ship. So instead, he opted to rush to get past
the Scillies and into Milford Haven in order to make
the transfer, thus increasing the pressure to make
good time.
The third human error in this chain was another
poor decision by the master. He decided, in order to
save time, to go through the Scilly Islands, instead of
around them as originally planned. He made this
decision even though he did not have a copy of the
Channel Pilot for that area, and even though he was
not very familiar with the area.
The final human error was an equipment design
error (made by the equipment manufacturer). The
steering selector switch was in the wrong position:
it had been left on autopilot. Unfortunately, the
design of the steering selector unit did not give any
indication of its setting at the helm.
So when the captain ordered a turn into the
western channel through the Scillies, the helmsman
dutifully turned the wheel, but nothing happened.
By the time they figured out the problem and got the
steering selector back on “manual”, it was too late to
make the turn, and the TORREY CANYON ran
The second case presented is the grounding of
EXXON VALDEZ. The ship’s compliment
consisted of four deck officers (captain, chief mate,
second mate and third mate), four engineering
officers, one radio electronics officer, six able-
bodies seamen, three unlicensed engine personnel
and two cook/stewards. The vessel personnel in the
deck department stood two four hour watches each
day with eight hours off in between. All other
personnel were day workers. According with
international minimum safety manning for this ship
would be fifteen crew members (Exxon Valdez had
20 crew members when she grounded on Bligh
Captain had been off the ship during the day she
was loading crude oil in Valdez port. Captain was
drinking that day. According blood analyses after
that his alcohol concentration was approximately
.285 at the time he boarded the ship, to do so without
showing some evidence of physical impairment or
needing some assistance. Additionally person
contacted after by the investigators reported none of
the EXXON VALDEZ crew members returning to
the vessel were under the influence of alcohol.
During the time the pilot was aboard the ship
Captain was off the bridge for approximately one
hour and thirty five minutes. The pilot smelled
alcohol on his breath.
Later on March 23rd, shortly prior to his relief,
the helmsman responded to an order from the master
to sail the ship 180
and put her on automatic pilot.
Helmsman was puzzled by this order. He didn’t
check it with the master. The master left the bridge
but not before asking the third mate, if he felt
comfortable sailing the ship under these conditions.
despite his limited experience in sailing the ship at
all, he replied that he did.
At 23:47 LT the ship left the Traffic Separation
Scheme going into the inbound lane to avoid the ice.
At 23:55 LT the helmsman was relieved. The ship
was on “load program up” which meant she was
increasing her speed while exiting the harbor. Thus,
EXXON VALDEZ was traveling at 12 knots and on
automatic pilot just prior to hitting Bligh Reef.
Putting the ship on automatic pilot in confined
waters and not telling the third mate the master had
done so was extremely inconsistent with normal
practice. At his relief, the helmsman reported to the
third mate that the ship was on automatic pilot,
something the third mate did not know about. The
third mate did not discuss the reason for the
automatic pilot with the master.
The third mate holds a second mate’s license, and
first sailed as the third mate on an Exxon tanker in
January, 1987. He had sailed on five tank vessels
owned by the company and had been employed by
Exxon for nine years. He had completed
approximately 18 voyages in and out of Valdez,
sailing in both unlicensed and licensed categories. At
the time of the grounding he had approximately 199
days of at sea experience as a third mate.
The night before he slept 6 hours, then after lunch
had a cat nap and relieved the chief mate for supper
and worked through to the grounding. The third mate
had only about a year’s experience as a deck officer.
The situation is further complicated because the
chief mate had worked the entire time of the loading,
was asleep, and was unavailable as an additional
resource. In addition to his bridge duties, the cargo
is the primary responsibility of a chief mate in
the Merchant Marine. This includes loading and
discharge of cargo could only be conducted by the
second and third mate on duty, the chief mate is
normally on hand when loading and discharging are
started and concluded. The ship left port at about
21:00 LT.
The third mate decided not to call his relief, the
second mate, until after they cleared the ice. The
third mate determined there was .9 mile between
Busby Island and the ice floe and felt he could pass
around the ice. The master left the bridge at 23:52
LT. the third mate relied considerably on the radar,
but did not correlate the radar information with the
navigation charts through position fixing. The
submerged reef was not displayed on the radar.
According with bridge organizational manual
used by Exxon, in this situation is stated that two
officer be on bridge during this transit. The chief
mate was sleeping. Some time before midnight the
third mate put the ship in hand steering condition. At
the same time he plotted the ship as 1.1 miles from
Busby Island. Before midnight the AB reported a red
light flashing
seconds to
the third mate.
He acknowledged her and stated that he knew the
light to be Bligh Reef. The third mate ordered a right
10 degree rudder but the vessel did not move to this
position. There is a six minute delay before the third
mate and helmsman respond to the fact that the ship
did not begin to turn.
About this time the AB reported the light flashing
every 4 seconds on the wrong side of the ship. Now
the third mate order a right 20 degree rudder.
Moving at 12 knots while the ship was still engaged
in maneuvering evolutions to avoid ice violated
prudent ship handling practices while increasing risk
of damage to the ship if ice floes had been struck. He
then orders hard right rudder.
When the ship hit the reef the third mate ordered
a hard left rudder to get the ship to stop swinging to
the right and prevent the stern from swinging
around. The ship had clearly skidded into Bligh
Reef. The helmsman was confused about some
aspects of the situation. He also reported that the
third mate was panicky. The chief engine stop the
engine at 00:20 LT.
For about 45 minutes the master tried to get the
ship off the reef, probably moving from dead slow
ahead to full ahead, and finally slowing down and
stop. The chief engineer had advised the master not
to move the ship. Vessel Traffic Service had advised
to move cautiously. The company declared that the
master was not trying to get the ship off the reef
because he never put the ship astern.
The chief mate was awakened by the grounding.
He went to the cargo control room to assess the
damage. He determined that the stress on the ship
exceeded acceptable limits and took this information
to the master. The chief mate performed further
analyses and concluded that if the vessel were not
supported by the reef it would capsize. He relayed
this information to the master who, for an additional
half hour tried to get the ship off the reef.
In this case, like in the first presented, a amount
of human errors concurred to the disaster. The errors
are from the area of navigation and ship characteristics
acknowledgement, bridge management regulations,
communication’s on bridge and on board the ship,
crew competencies.
Below we’ll try to show some of these errors,
according with facts and regulations applied for this
A number of dynamics occurred on the bridge.
The first is that the two key players aren’t there. The
company manual stated that the master or chief mate
must be on the bridge while exiting port and the law
requires a first class pilot’s license or endorsement
for the waters. The situation warrants the added
responsibility of the master to be on the bridge, not
the chief mate during loading and discharging
operations. Sufficient redundancy might have been
in the system if one of these people had acted as a
second pair of eyes for the third mate. Second, no
one checked the reasoning behind orders. From this
account we don’t know if the helmsman may have
had reason to question the situation (gyro, load
program up conditions). The AB may have
questioned in his mind what they were doing. If he
did he didn’t find a way to direct attention to that
question without putting himself in danger of
incurring the third mate’s wrath.
Overall, one might suspect this kind of
unprofessional seamanship on the part of the captain,
the third mate and the helmsman had occurred
before. Such behaviors usually don’t emerge full
blown, they grow over time. There is sufficient
evidence from the company that the captain had
problems managing people and there is some similar
evidence that the third mate found it difficult to keep
supervisors informed about what he was doing.
There is nothing that indicates training or a culture
that values open communication among bridge
personnel. An appropriate culture of safety and
vigilance seems not only to have been in place. The
watch cycles (4 on, 8 off, 4 on, daily) seems an
inherent part of the organization.
Looking at the performance evaluations of the
helmsman, it is clear he was not very competent. A
master should not leave the team of an incompetent
helmsman and a third mate with little experience to
run a tanker through an ice field. In this case the
pulls and pushes on the master lead to his failing to
think about this issue.
The EXXON VALDEZ didn’t operate in isolation
from the relationships various participants had with
one another. The pilot smelled liquor on the master’s
breath and didn’t report it to anyone. The
relationship between pilots and master’s is sensitive,
and the pilot’s job future is an important respects
depends on what the master thinks of him. Through
this relationship seems cast in stone it may well be
time to examine it thoroughly. Similarly, the
relationship of the VTS at Valdez and the EXXON
VALDEZ was one of very little attention even to the
giving of advice. This kind of quasi advice only
versus direction issue must be looked at in both
Crew aren’t in place for the operation of a culture
which stresses the existence of risk and risk
avoidance. They aren’t in place for good
communication among the parties, they may not be
in place for engaging in good training which can
help the bridge team interact appropriately. In
addition, if anyone in the bridge group was not
competent, the rewards are not in place for getting
rid of that person or retraining him.
We have seen that human error (and usually multiple
errors made by multiple people) contributes to the
vast majority (75-96%) of marine casualties, making
the prevention of human error of paramount
importance if we wish to reduce the number and
severity of maritime accidents. Many types of human
errors were described, the majority of which were
shown not to be the “fault” of the human operator.
Rather, most of these errors tend to occur as a
result of technologies, work environments, and
organizational factors which do not sufficiently
consider the abilities and limitations of the people
who must interact with them, thus “setting up”
the human operator for failure. Human errors can
be reduced significantly. Other industries have
shown that human error can be controlled through
human-centered design. By keeping the human
operator uppermost in our minds, we can design
technologies, work environments, and organizations
which support the human operator and foster
improved performance and fewer accidents.
McCallum M.C., Raby M. & Rothblum A.M. 1996 Procedures
for Investigating and Reporting Human Factors and Fatigue
Contributions to Marine Casualties. U.S. Coast Guard
Report No. CG-D-09-97. AD-A323392.
Perrow C. 1984 Normal Accidents: Living with High-Risk
Wagenaar W.A. & Groeneweg J. 1987 Accidents at sea:
Multiple causes and impossible consequences.
Hanzu-Pazara R. & Bocanete P. 2006 The influence of team
error in maritime industry, Transport Book of Constanta
Maritime University Navigation Dept, no. 4.
Reason J, 1990 Human errors.