Human Factors (NG4S245)

Assignment 1
Charlie Jenkins (12010413)
Supervisor: Bethan Llewellyn

Case Study: Boeing 767-200, N799AX, 28 June 2008SanFrancisco

Executive Summary

Charlie Jenkins (12010413) University of South Wales

Human Factors

This report is primarily concerning an analysis of an incident which occurred on the

28th of June 2008 at San Francisco airport. This incident involved a ground fire in a Boeing
767-200, N799AX cargo aircraft, which was operating as flight 1611. The aircraft was
operating under 14 Code of Federal Regulations Part 121. This is essentially an air carrier
certification process that ensures the operator is implementing the relevant safety
procedures and regulations and is also able to manage hazards in the operating
environment appropriately. The primary cause of this incident as determined by the NTSB
report AAR-09/04 (NTSB, 2009) was the design of the supplemental oxygen and the lack of
sufficient separation between the electrical wiring system and the electrically conductive
components of the oxygen system. The minimal separation allowed a short circuit to
penetrate a combustible oxygen hose, causing a fire in the supernumerary section of the
aircraft, this fire then rapidly spread to surrounding areas. An external contributing factor to
the incident was the FAAs (Federal Aviation Authority) failure to require the installation of
nonconductive oxygen hoses after safety issues were raised by Boeing and ABX Airs lack of
action on an oxygen system that clearly showed consistent leaks and continuous issues.
As this analysis is constrained mainly to the issues that can relate to Human Factors,
very specific areas were discussed. These areas are not only constrained to an individual,
as one would assume with most Human Factors issues, but also to organisational omissions
and behavioural aspects, which all contributed towards the incident occurring. To identify and
highlight all of these issues, the swiss-cheese model and the error chain model were
adopted and used in conjunction with the relevant standards. This will be included in the
appendices. What became very clear throughout the report is that there were also
organisational cultural issues which led to the maintenance individual failing to address the
oxygen system discrepancies appropriately. It also seemed that FAA and ABX failed to
recognise the safety issues in relation to the oxygen system.

List of Contents
Executive Summary..2
Page 2 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

List of Contents.3
List of Abbreviations.4
List of Figures, Tables and Graphs5
Chapter 1: Introduction6
1.2: Incident Description.6
1.3: Contributing Factors8
Chapter 2: Findings...11
2.1: Organisational omissions/failures...11
2.2: Individual omissions/failures13
Chapter 3: Conclusion and Recommendations14
Appendices..15
Bibliography.17

List of Abbreviations

Page 3 of 18

Charlie Jenkins (12010413) University of South Wales

CAMP

Human Factors

Continuous Airworthiness Maintenance Program

ARFF

Aircraft Rescue &

Firefighters
HRET

High Reach Extendable Turret

SPN

Skin Penetrating Nozzle

FAA

Federal Aviation Authority

CAA

Civil Aviation Authority

AD

Airworthiness Directive

SB

Service Bulletin

CASP

Continuous Analysis & Surveillance Program

List of Figures, Tables and Graphs

1.1: Forward Layout of the incident Aircraft (NTSB, 2009)
Page 4 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

1.2: Incident Timeline (NTSB, 2009)

1.3: Oxygen Supply Hose (NTSB, 2009)
1.4: Electrical wiring in close proximity to oxygen system (NTSB, 2009)

Chapter 1: Introduction
In this report, a case study that has several Human Factors related issues will be
analysed, with prominent issues highlighted and discussed. Specifically, the limitation of
Page 5 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

human performance in the workplace in the respect of minimising and efficient methods of
managing errors. As most of the fact finding from the investigation was carried out by the
NTSB, this will be the primary source of information (NTSB, 2009).The specific case study
that will be used in this assignment is the incident is as follows: Aircraft: 767-200,
Identification Number: N799AX, Date: 28 June 2008, Location: San- Francisco. A majority of
information for this report was taken from the NTSB report (NTSB, 2009)
To understand the individual and organizational failures, it is important to understand
how airworthiness and continuous safety is achieved. When a fault is pinpointed, a
manufacturer publishes a Service Bulletin, which recommends the maintenance or a specific
part to be replaced. However, these SBs are not compulsory, they only become compulsory
when the relevant regulating authority (FAA, CAA etc) formulate an AD (Airworthiness
Directive).
Incident Summary: in this particular incident, several human and organizational
errors lead to a potentially catastrophic ground fire which occurred before engine start up.
Although the crew escaped unharmed via the cockpit window emergency escapes, there
was substantial damage to the aircraft. Several procedural flaws and other aspects were
contributory factors to the incident.
1.2: Incident description
Whilst the pilots were doing ground checks and preparing for take-off, the aircraft
N799AX experienced a ground fire before engine start up. The captain and first officer
managed to safely evacuate the aircraft through the cockpit window emergency exit and
were unharmed. However, the aircraft was substantially damaged. Whilst carrying out all of
the engine start checks, the pilots heard pop and hissing noises, the first officer went to
inspect the supernumerary compartment and saw that a fire had started near the right end of
an enclosure that contained components of the supernumerary supplemental oxygen system
(see Figure 1.1). After identifying that there was a fire, the flight crew contacted the ground
controller to inform that the Aircraft Rescue and Firefighting (ARFF) was needed because of
the cargo fire. They then proceeded to carry out the fire and evacuation checklist and
successfully evacuated the aircraft via the cockpit window exits, this was due to the severity
and close proximity of the fire blocking and sealing the primary exit doors. The ARFF then
extinguished the fire. You can see the incident timeline simplified in Figure 1.2.

Page 6 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Figure 1.1. Forward layout of the aircraft (NTSB, 2009)

1.3:

Figure 1.2. Accident Timeline (NTSB, 2009)

Contributing Factors

Page 7 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

After the incident, an extensive NTSB report was done in order to establish the root
causes. According to this report, the following pertinent factors contributed towards the
incident:

Oxygen hoses were manufactured using electrically conductive material.

The general composition of the oxygen supply hoses is as seen in Figure 1.3. A rubber
internal hose with stainless steel sleeve and a metal internal spring to ensure that supply of
oxygen wasnt faltered and the integrity of the hose was maintained. As the spring makes
contact with a metal surface which is in close proximity of electrical supply wires which could
electrically charge the spring, the rubber hose could also act as a fuel, this presented a clear
contributing factor.

Figure 1.3. Oxygen Supply Hose (NTSB, 2009)

Electrical cables, which were close to the oxygen supplies, didnt have
satisfactory positive separation.

As part of the NTSB investigation, similar Boeing 767 airplanes that were owned by ABX had
to be inspected as a safety measure but also to aid in the fact finding process and evaluating
the probable cause of the incident. During the inspections, it was found that electrical wiring
was routed near the oxygen supply tubing (see Figure 1.4). This poses a clear risk and
NTSB state that this was the most likely cause of the incident.

Page 8 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Figure 1.4. Electrical wiring in close proximity to oxygen system (NTSB, 2009)

Airworthiness Directive Process

Boeing received reports that electrical energy had caused leaks in the oxygen system that
were installed on their airplanes, after this they released a Service Bulletin (SB) (September
1999), which recommended the replacement of existing oxygen hoses in the cockpit with
nonconductive hoses. The FAA failed to acknowledge the potential impacts of this issue and
didnt require compliance to this SB by not implementing an Airworthiness Directive (AD)
until after the accident.

Electrical Grounding of Oxygen System

Inspection of the supplemental oxygen system found that although it was grounded through
the rubber clamps, the oxygen hose and the fittings themselves were not evaluated as
becoming a potential ground path for the electrical wiring that was passing in close proximity.
Investigations also determined that the path of least resistance was through the hoses
pictured above. The hoses are also susceptible

Aging Oxygen Hoses

The investigation also concluded that although the oxygen hoses on the accident airplane
were consumed by the fire, no life limit was set for them. Some hoses had been in service
for more than 40 years. Older hoses on other 767 aircraft owned by ABX Air were much
stiffer than newer ones. Stiffer hoses are clearly more likely to crack and enable oxygen
leaks, facilitating a fire. The PVC hose material is also susceptible to degradation over time
by means of temperature, light and items used in regular aircraft maintenance.
Page 9 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Reading Lights potentially became ignition source.

As most cargo aircraft are just conversions from commercial airlines, a majority still have
reading lights installed throughout the cargo compartment. It was found that the PSU
(Passenger Service Units) assemblies were installed on ABX Airs 767 aircraft in the
supernumerary compartment in close proximity to the supplemental oxygen system. It was
found that the reading light could be rotated so that the grounded socket housings could
touch the electrical contacts for the PSU switches. The manufacturer did provide rubber
boots, which acted as an isolation from the switch assembly. However, these were found to
not be installed on a majority of the aircraft, apparently due to the difficulty of installation.

Smoke Detection System

As the aircraft in this incident was converted from a commercial aircraft, there was no smoke
detection system installed in the supernumerary compartment. This is due to the fact that
operators inherently rely on passengers and flight crew to report any incidents. However, it is
a requirement to have smoke detectors in the lavatory and cargo compartments. Both of
which were installed but due to the lavatory doors primarily being closed, the detection of
smoke was significantly delayed.

ABX Airs Continuing Analysis Surveillance Program

ABX Air had a continuous airworthiness maintenance program (CAMP) tracker in place
which was labelled as CASP (Continuing Analysis and Surveillance Program). The purpose
of this program was to ensure quality reliability. In order to do this, the program used trend
analysis to pinpoint repetitive airplane discrepancies and provided notices. The scope of this
analysis was to provide a notice if anything was generated within specific time periods. Such
as: three defects within 15 days, five defects within 30 days or one or more defects within 2
days of a notice closing date. ABX then held weekly reliability meetings to assess any
repetitive discrepancies identified by the advisory notices and take any action that was
relevant. Despite all of this being in place, ABX Air failed to act upon four advisory notices
which were produced within 7 months on the accident aircraft. This is due to the meetings
only considering 3 months prior.

Page 10 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Chapter 2: Findings
2.1: Organisational Omissions/Failures
Throughout the NTSB report, it is clear that there is several Organisational Issues that need
to be addressed and managed. These issues, if addressed as they should have been, would
have significantly reduced the risk of this incident occurring.
To start, there seemed to be a clear ignorance by the FAA of previous issues that had
been raised by the NTSB that could have reduced the risk of this issue occurring. This can
be demonstrated by Safety Recommendations A-98-1 and -2, which mentioned the
following:
Review the design, manufacturing, and inspection procedures of aircraft
manufacturers, and require revisions, as necessary, to ensure that adequate clearance is
specified around electrical wiring, in accordance with published FAA guidelines. (A-98-1)
(NTSB, 2009)

Review the existing designs of all transport-category airplanes to determine if

adequate clearance is provided around electrical wiring, in accordance with published
FAA guidelines. If deviations are found, require that modifications be made to ensure
adequate clearance. (A-98-2) (NTSB, 2009)
Although the FAA did implement AC 43.13-1A which requires 0.5 inches separation,
this didnt seem to be sufficient enough to prevent the risk of this incident occurring. It is the
authors belief that the FAA didnt recognise the fact that electrical arcing can still easily
occur with a separation gap of 0.5 inches. It is also true that this AC is only advisory and
therefore not always abided. Making this separation a requirement should have been made
a priority.
The FAA also failed to recognise the risk of the oxygen system itself conducting
electricity and therefore didnt require any prevention methods (such as grounding straps) to
be implemented. This implementation could have significantly reduces the probability of fire
occurring. Paired with this, the FAA failed to recognise the ignition capability of the oxygen
hoses, and prevention of conductive material being used was not apparent.

Page 11 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

The FAA also failed to recognise the potential of passenger reading lights becoming
an ignition source, with rubber boots being provided for electrical grounding but the
installation of these not being a requirement and also being very time consuming. The time
consuming obviously means significant cost impacts to the operator, therefore a majority of
the rubber boots were missing.
Boeing actually highlighted the potential of conductivity in oxygen hoses and
promptly produced an SB, however, the FAA failed to recognise the significance of this issue
and didnt release an AD. This implementation would have more than likely avoided the
incident all together.
Paired with the failures that seemed evident in FAAs processes, there was also
process flaws in ABX Airs organisational procedures. Every operator is required to
implement a CAMP, ABX Airs version of this was the CASP. Regardless of the fact that ABX
Air did have a program in play, they failed to address the incident airplanes oxygen system;
which showed excessive discrepancies and displayed clear signs of a chronic problem. This
was primarily due to the fact that the scope for CASPs consideration was only 3 months.
ABX Air also didnt take notice of Boeings SB due to this 3 month period. Had their
surveillance period been extended, they would have seen 28 discrepancies and the SB
which was produced by Boeing.
The severity and longevity of the incident could also be partly blamed on the FAA. In
order to extinguish the fire, the ARFF had to use a HRET and SPN, this was due to the door
mechanisms being disabled by the fire. These particular tools are notoriously difficult to use
effectively. The NTSB had previously released a safety recommendation A-07-100 (2007),
which stated the following:
Provide guidance to aircraft rescue and firefighting personnel on the best training
methods to obtain and maintain proficiency with the high-reach extendable turret with skinpenetrating nozzle. (NTSB, 2009)
Although the FAA did issue a CertAlert to recommend training, they failed to highlight
the importance and relevancy of such training. This demonstrates the FAAs lack of urgency
with the matter. Further to this, the FAA only issues an AC (Advisory Circular) in August
2009, a long time after the incident.
Page 12 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

2.2: Individual Omissions/Failures

In the previous few months running up to the incident, (January 2007 June 2008),
the oxygen system was serviced 50 times for routine maintenance of the oxygen system
and as a resultant of reported discrepancies. The NTSB report shows that most of the 50
oxygen services occurred after the aircrafts last C check which was on July 23, 2007.
Maintenance records clearly show that, after this event, there was a frequent refill of oxygen
bottles required. This begs the question as to whether any damage was done to a system
during the C check in July. This could easily have happened and not been reported or even
noticed.
Despite this, it should have been apparent not only to the organisation, but also to
the individuals that the frequency of oxygen system servicing and refills was depicting a
system with a chronic leak. Concurrently with this issue, it also became apparent that some
of the maintenance personnel didnt always record oxygen system servicing in the
maintenance logbooks. The cause of this was probably due to the maintenance personnels
lack of understanding of the potential hazard that accompanies an oxygen system but could
also be due to work related pressures; such as time constraints.
One of the previously mentioned unreported incidents actually occurred on the
predeparture check of the accident aircraft. The mechanic removed the supernumerary
oxygen supply bottle from the aircraft due to it reading below the AMM (Aircraft Maintenance
Manual) requirements. The mechanic then reinstalled after trying to add oxygen. The
mechanic stated that he didnt add this work to the logbook as no maintenance was
conducted. The mechanic failed to check the bottle to ensure that it contained usable
oxygen available for use and also failed to realise the importance of logging this work on a
safety-critical system.
There was also a significant delay in the extinguishing of the fire (43 minutes after the
first ARFF vehicle arrived). This was partly due to lack of sufficient training, and on the FAAs
back. Despite this, the author feels there was still some individual failures in terms of
communication between the firefighters to ensure efficiency. The driver of the HRET/SPN
inserted the SPN through the cockpit window and attempted to spray the extinguishing agent
towards the fire. Due to lack of communication this persisted for a prolonged period,

Page 13 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

meaning this contributed towards the extended blaze period and extensive damage to the
aircraft.

Chapter 3: Conclusion and Recommendations

In conclusion, it seems evident that there were not only individual failures, but also
significant organisational failures which were all contributory factors to the incident. It seems
apparent that the most significant amount of factors seemed to stem from organisational
issues rather than individual. This highlights a contrast of popular belief that usually a
singular person is to blame for human error.
It is the authors belief that the primary cause for the reluctance progression of FAAs
AD process is not only due to there being a lack of understanding the importance and safety
concerns surrounding the oxygen system, but also due to logistics and monetary concerns.
Although safety should really be a priority, it still seems that a majority of companies
consider overheads and profit before they consider overhauling a system due to safety
concerns. The enforcement of a proactive approach to maintenance should be
recommended by the FAA, because it seems that operators simply wait for an AD rather
than follow the SB recommendations. The fact that the manufacturer was producing oxygen
hoses with conductive materials, shows the clear disregard for any potential issues, again,
cost being the prominent factor here.
In addition to this, if there was an efficient SMS (Safety Management System), paired
with an EMS (Error Management System) then the probability of incidents such as the afore
mentioned 767 could be significantly reduced. As is stated in CAP715;
Within a maintenance organisation, data on errors, incidents and accidents should
be captured with a Safety Management System (SMS), which should provide mechanisms
for identifying potential weak spots and error-prone activities or situations. Output from this
should guide local training, company procedures, the introduction of new defences, or the
modification of existing defences. (Civil Aviation Authority , 2002)

Page 14 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Human error is a significant problem in aviation, but is completely unavoidable. The

only action that an organisation can take is a proactive approach to error and safety
management.

Mitigation

is

key.

This

implementation

is

therefore

the

authors

recommendation.

Page 15 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Appendices

Organisation
failed
to install
grounding
boots
on passenger
reading lights due to logistical and time
All of these holes in the defences aligning caused
the error
to occur
within
the limited
window
of opportunity

Failure to sufficiently electrically ground oxygen systems and have sufficient separation minima. Composition of oxygen hoses a

Substandard practices of operators. Maintenance engineer failed to understand the safety importance of an integral and pressure conform

ABX Airs poor CAMP and insufficient scope in terms of the CASP. Failure to recognise continuous and concurrent problems with oxygen s

Organisational failures, in terms of the FAAs failure to recognise the importance of Boeings recommendation through the SB that they issued. Failure to

Appendices 1: Swiss Cheese Model (Reason, 1990)

Page 16 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Management

Governing Body

Error Chain Model: A lot of incidents compromise of many human factors making a chain of errors. If any link in this chain had been broken by buildi

Append
Page 17 of 18

Charlie Jenkins (12010413) University of South Wales

Human Factors

Bibliography
Airbus, 2005. Human Performance Error Management. [Online]
Available at:
http://www.airbus.com/fileadmin/media_gallery/files/safety_library_items/AirbusSafetyLib_FLT_OPS-HUM_PER-SEQ07.pdf
[Accessed 16 January 2016].
Beoing, 1991. Accident Prevention Strategies: Commercial Jet Aircraft Accidents World
Wide Operations. 1 ed. s.l.:Boeing.
Civil Aviation Authority , 2002. An Introduction to Aircraft Maintenance Engineering Human
Factors CAP 715, JAR 66: CAA.
Civil Aviation Authority, 2002. Maintenance Error Management Systems. [Online]
Available at: http://www.raes-hfg.com/reports/23oct02-fatigue/23oct02-mems.pdf
[Accessed 16 January 2016].
Ministry of Defense, 2014. Defense aviation error management system (DAEMS). [Online]
Available at: https://www.gov.uk/government/publications/defence-aviation-errormanagement-system-daems-documentation
[Accessed 15 January 2016].
NTSB, 2009. AAR-09/04. [Online]
Available at: http://www.ntsb.gov/investigations/AccidentReports/Reports/AAR0904.pdf
[Accessed 02 January 2016].
Reason, J., 1990. Human Error. 1st ed. New York: Cambridge University Press.

Page 18 of 18