Aviation Safety Investigation Report

200305496

The effectiveness of the fire fighting

arrangements for Bankstown Airport, as they
affected transport safety on 11 November 2003

1
FACTUAL INFORMATION

Accident sequence
On 11 November 2003, a Piper Seneca, registered VH-CTT, was seen to diverge right
during a go around from runway 11 Right at Bankstown Airport, NSW. It maintained
a constant height above the ground and; when about halfway along the runway, the
aircraft’s nose lifted and it banked steeply to the right before impacting the ground in
a near vertical nose-down attitude. A student pilot in the left seat and an instructor
pilot in the right seat were the only occupants.
A fire commenced when the aircraft impacted the ground or shortly after the impact.
The fire intensified after the aircraft came to rest. The main cabin door, located over
the right wing, separated from the aircraft during the accident. The instructor pilot
vacated the aircraft through that opening about 30 seconds after the aircraft came to
rest. The student pilot was fatally injured. The instructor pilot received severe burns
and was treated in hospital for three and a half weeks before succumbing to those
injuries.

Figure 1: Post-impact dust and fire

Effectiveness of fire fighting services

On 1 December 2003, the Minister for Transport and Regional Services signed an
Instrument of Direction to the Australian Transport Safety Bureau (ATSB). That
instrument directed the ATSB to ‘investigate the effectiveness of the fire fighting
arrangements for Bankstown Airport, as they affected transport safety at Bankstown
Airport on 11 November 2003’. Effectiveness is defined as ‘serving to effect the
purpose; producing the intended or expected result’. 1 The instrument was issued to
the ATSB on 15 December 2003.

1
Macquarie Dictionary (revised third edition), The Macquarie Library, NSW, 2003

2
International requirements
Australia is a signatory to the Convention on International Civil Aviation (Chicago
Convention). This agreement sets out standards and recommended practices (SARPs)
for signatory States. However, Australia has notified the International Civil Aviation
Organization (ICAO) of a difference from those SARPs for national rescue and fire
fighting services. The difference states that ‘rescue and fire fighting services to Annex
standards are not available at some international and international alternative
aerodromes’. Bankstown Airport is not an international or an international alternative
aerodrome.
The ICAO Airport Services Manual, Part 1, Rescue and Fire Fighting (3rd Edition
1990) states in Chapter 2, Paragraph 2.7.1 that the ‘operational objective of a rescue
and fire fighting service should be to achieve response times of 2 minutes and not
exceeding 3 minutes to the end of each runway in optimum conditions of visibility
and surface conditions, as well as any other part of the movement area’.
Paragraph 2.7.2 states that ‘to ensure continuous agent application at the appropriate
rate, any other rescue and fire fighting vehicles required to carry the amounts of
extinguishing agents specified in [airport category2] Tables 2-2 and 2-3 [of Paragraph
2.3] should respond no more than 60 seconds after the first responding vehicle’.
Paragraph 2.3, titled ‘Amounts of Extinguishing Agents’, lists these water, foam and
complementary agent specifications and the discharge rates.
Table 2-2 details the minimum usable amounts of extinguishing agents for aqueous
film forming or fluoroprotein foam. Table 2-3 details the minimum usable amounts of
extinguishing agents for protein foam. Paragraph 2.6.1 states that ‘the quantities of the
various extinguishing agents to be provided in the rescue and fire fighting vehicles
should be in accordance with the airport category tables’.
Based on the 2002 Bankstown Airport aircraft movement data, for the purposes of
aerodrome rescue and fire fighting services (ARFFS), Bankstown Airport would have
been a Category 3 aerodrome.3 That category aerodrome would normally require a
minimum of one rescue and fire fighting vehicle.4
The ICAO standard for aerodrome rescue and fire fighting services is based on the
following statistics and parameters:
- about 70 per cent of aircraft crashes occur on aerodromes
- of those that occur on aerodromes, 90 per cent are survivable
- people on board a major aircraft that is involved in fire can survive up to four
minutes

2
The airport category for ARFFS should be based on the dimensions of the aeroplanes using the
airport as adjusted for their frequency of operations.
3
Assuming Hawker Siddley 748 (MAUW 20200 kg) as largest aircraft using the aerodrome but with
less than 700 movements per month, the category can be reduced by 1 (from Cat 4 to Cat 3).
4
ICAO Annex 14 – Aerodromes Chapter 9.2.23 recommends a minimum of one vehicle for an
aerodrome classified Category 1 to 4.

3
- intervention of an aerodrome rescue and fire fighting service within the four
minutes can extend that time limit allowing people on board to be rescued.5

National regulations
Safety regulation of civil air operations in Australia and the operation of Australian
aircraft overseas is the primary function of the Civil Aviation Safety Authority
(CASA).
In the early 1990s a decision was made to remove ARFFS at some aerodromes.
Services were removed from all general aviation airports (GAAP)6 following
consultation with the aviation industry and the aerodrome operators. Rescue and fire
fighting services were withdrawn from Bankstown in 1991.
CASA safety regulations in relation to ARFFS are contained in Civil Aviation Safety
Regulation (CASR) Part 139 Subpart H - Aerodrome rescue and fire fighting service.
CASR Subpart 139 H was adopted on 26 June 2002, for commencement on 1 May
2003, following industry consultation conducted since March 2000.7
It was intended that Part 139 Subpart H would place an obligation on aerodrome
operators to provide an ARFFS, for aerodromes from, or to, which an international
passenger air service operated, and any domestic aerodrome through which more than
350,000 passengers passed on air transport flights during the previous financial year.
This standard was aimed at minimising the risk for the greatest number of passengers.
However, subsequent actions following the tabling of a disallowance motion in
Parliament removed both the obligation for anyone to provide an ARFFS or requiring
the provision of an ARFFS. The Department of Transport and Regional Services
(DOTARS) has taken action to re-insert establishment criteria in the CASRs.
Of Australia’s 600 aerodromes, Airservices Australia provides ARFFS at eight capital
city and nine regional aerodromes8 in line with the Government’s policy of providing
ARFFS at aerodromes exceeding 350,000 passengers in the previous financial year, or
that have international air services.9 There is nothing precluding an airport operator of
an airport that falls outside the criteria, should they choose to, from providing a non-
ARFF service at their airports, or from requesting a certified supplier to provide an
ARFFS to an appropriate standard.
The Manual of Standards Part 139 H – Standards Applicable to the provision of
Aerodrome Rescue and Fire Fighting Services, states that all operational ARFFS staff
must comply with CASA standards that require a current qualification and certificate
of competency commensurate with the functional role at a specific location.

5
House of Representatives Standing Committee on Transport and Regional Services report, Regional
Aviation and Island Transport Services: Making Ends Meet, December 2003.
6
Includes Parafield SA, Jandakot WA, Camden NSW, Moorabbin Vic., and Archerfield Qld.
7
Information about the consultation process is available from the CASA website
www.casa.gov.au/avreg/newrules/casr/139h.htm
8
Airservices Australia data available at www.airservicesaustralia.com/services/ps7/contracts2.asp
9
House of Representatives Standing Committee on Transport and Regional Services report, Regional
Aviation and Island Transport Services: Making Ends Meet. The report is available at
www.aph.gov.au/house/committee/trs/aviation/report/contents.html

4
Those competencies include casualty assistance, emergency care, emergency life
support techniques and how to operate life-support equipment.10
The Department of Transport and Regional Services collects aviation statistics on
passenger movements at Australian aerodromes. However, the data gathered is limited
to passengers from international, domestic and regional airline scheduled air transport
flights at 94 aerodromes. That data did not include passengers carried on charter or
other non-scheduled air transport flights, and it did not include those aerodromes that
received less than 7,000 revenue passenger movements in 2002–03. Bankstown
Airport does not have scheduled air transport passenger flights and is not an
international airport.11 The passenger data from Bankstown and other GAAP
aerodromes is excluded from the DOTARS annual statistical publication.
The Aeronautical Information Publication En Route Supplement Australia (ERSA)
details the ARFFS for each aerodrome listed. The Bankstown ERSA entry has no
ARFFS detailed.

Parliamentary activities
On Monday 1 December 2003, the House of Representatives Standing Committee on
Transport and Regional Services tabled its report Regional Aviation and Island
Transport Services: Making Ends Meet. The report follows an inquiry into
commercial regional aviation services in Australia and transport links to major
populated islands.
The inquiry looked at the cost of aviation [aerodrome] rescue and fire fighting
services, and the conditions under which these services are now provided following
industry concerns. It found that there was some disagreement in the evidence about
the justification for fire fighting and rescue services being provided at aerodromes.
The cost of providing the service is determined by the equipment needed to meet the
standards for delivering water and foam within an aerodrome during a call-out. At
some aerodromes around the world, and in Australia, it has been found to be more
cost effective for aerodrome fire fighting and rescue services to be provided by the
fire service serving the local community. At Bankstown, where there are no scheduled
passenger services, the aerodrome fire fighting and rescue service is provided by the
New South Wales Fire Brigades.
The inquiry found that there were two issues to be considered:
- the provision of rescue and fire fighting services to aerodromes with limited
numbers of passenger landings
- whether these services are provided by Airservices Australia or by local fire
services.
The committee report contained the following recommendations:
- The Department of Transport and Regional Services and Airservices Australia
introduce a universal service charge for aerodrome rescue and fire fighting

10
Certificate II and III level of Australian Fire Competency.
11
Data on the number of passengers using non-scheduled services such as charter flights is not
available, as indicated by the preceding discussion.

5
 services at regional aerodromes to reduce the wide disparity in the charges for
those services and to reduce the overall impact of the charges on regional aviation
costs;
- The Department of Transport and Regional Services and Airservices Australia
form a working group with key stakeholders (such as the relevant local
government associations, town planning and standards bodies) to advise on the
strategic and optimal co-location of fire fighting services; and
- Airservices Australia provide the initial aerodrome rescue and fire fighting
equipment and crew training, at no cost, to communities where fire fighting
services become co-located.
The Government is currently considering its response to this comprehensive report.

Air traffic control response to the emergency

On the day of the accident involving VH-CTT, the air traffic control tower was
staffed by two controllers, one occupying the aerodrome control (ADC) position and
the other at the tower coordinator (COORD) position.12 Both controllers saw the
Seneca go around and diverge from the runway. When it was apparent that there was
likely to be an accident they commenced emergency response procedures.
The ADC activated the crash alarm at 1247:25 Eastern Summer Time, however he
incorrectly selected the console power switch. The accident was estimated to have
occurred at 1247:30. The ADC re-selected the console power switch and power was
restored at 1247:50. At 1247:55, the Bankstown Tower Coordinator/Surface
Movement controller notified the New South Wales Fire Brigades of the accident.
The ADC worked at both Camden and Bankstown control towers. The location of the
crash alarm and console power switch on the consoles were transposed in the
respective towers.
Airservices Australia advised that, where an ARFFS is provided, that service is
required to monitor landings and takeoffs from a fire control centre (FCC). A crash
alarm can be activated from an FCC that may result in an enhanced response to an
accident. Airservices also advised that ARFFS response times can be significantly less
than required in the CASRs at smaller aerodromes similar to Bankstown.

Personnel response to the VH-CTT accident

Very soon after the instructor had vacated the aircraft, personnel working near the
museum (located within the aerodrome), and from a factory complex adjoining the
aerodrome, attacked the fire using hand-held fire extinguishers. One person from the
factory, who was trained in fire fighting as part of the factory’s emergency response
team (ERT), administered first aid to the instructor and directed other staff from the
factory to lay out a fire hose from the ERT equipment trailer. At least 12 fire
extinguishers and the hose were used on the aircraft fire before the fire brigade
arrived.

12
The controller at the coordinator position was managing both coordinator and surface movement
control functions.

6
Figure 2: Aircraft accident site

New South Wales Fire Brigades (Brigades)

The Brigades’ objective in an emergency (24 hours a day, 7 days a week), in the
greater Sydney area, which included Bankstown Airport, was to attend the site of an
emergency incident within 10 minutes from the time of a call on 90% of occasions.
Fire fighting units from Bankstown, Revesby and Chester Hill were dispatched at
1248:53 with a rescue unit from Liverpool at 1249:19.
At 1256:49 the Bankstown Urban Pumper unit arrived at the aerodrome. That unit
was delayed in gaining entry to the aerodrome as the safety officer, responsible for
managing access to the airside of the aerodrome, was escorting an ambulance to the
accident site as the unit arrived at the assembly point. Consequently, the unit was
delayed outside the fence securing the aerodrome movement area. It was normal
practice for a safety officer to escort non-aerodrome personnel and vehicles onto the
aerodrome movement areas. At 1258, the crew reported by radio that they were,
‘…making way to the crash site’.
That first unit, at the accident site, was followed by other units at 1258:02 (the
Revesby Rescue Monitor unit), at 1258:56 (Chester Hill Urban Pumper unit) and at
1300:34 (Liverpool Rescue unit).
The Brigades assumed control of the site, extinguished the fire and requested
aerodrome and local industry personnel who had commenced to fight the fire with
portable fire extinguishers and the 38 mm fire hose, to vacate the site for their own
safety. Only two of the four units at the scene were used. The units returned to their
respective stations at 1417:00.
Data provided by the Brigades indicates that 9 minutes is the average time for a unit
to travel from Bankstown, Revesby, Chester Hill or Liverpool stations to the centre of
Bankstown Airport based on the Brigades’ planned road speed of 42 kph.
During 2001–02 and 2002–03, the Brigade responded within 10 minutes to 86% and
88% of calls within Bankstown Airport respectively. Of the 22 calls in 2001–02, two
were in response to an actual fire/explosion and of the 26 calls in 2002–03, three were
in response to an actual fire/explosion. The average response time in 2001–02 was 8
minutes 18 seconds and in 2002–03 it was 7 minutes 18 seconds.

7
Brigades’ data for the responses to the 1,992 calls during 2002–03 within the suburbs
of Bankstown, Revesby, Chester Hill or Liverpool (exclusive of aerodrome calls)
indicate an average response time of between 5 and 7 minutes.

Transport mode safety comparisons

In 2002, the ATSB issued a discussion paper on Cross Modal Safety Comparisons.13
The purpose of the paper was to look at the relative safety of various transport modes.
The study found that while relative safety rates between different transport modes are
difficult to compare, based on the best available data, high capacity regular public
transport air travel (involving transport category aircraft) is the safest form of
transport, while general aviation is significantly less safe than car travel.

ATSB accident database

A review of the Bureau’s aircraft accident database14 between 1980 and 2003 was
conducted to:
- ascertain the significance of the withdrawal of ARFFS from general aviation airports
in 1991
- establish the number and location of fire related fatal general aviation (GA)
accidents during the period.

Figure 3: Fire related fatal general aviation aircraft accidents per 100,000
flying hours
2.50

2.00

1.50
Rate

1.00

0.50

0.00
80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

00

01

02

03
19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

20

20

20

20

Year

Accident rate Fire related accident rate

13
Available at www.atsb.gov.au/public/discuss/cross_modal.cfm.
14
The review excluded non-powered general aviation accidents (gliding, parachuting and paragliding)
that would not be subject to in flight or post impact fire.

8
The data revealed that there were 638 fatal GA aircraft accidents and of those, there
were 106 in which the aircraft and the occupants were involved in a post-impact fire.
The data was normalised and a rate per 100,000 general aviation flying hours was
calculated to enable a comparison of the number of fire and non-fire related accidents.
The comparison revealed that there was no statistically significant change in the fire
related accident rate following the withdrawal of the ARFFS from general aviation
aerodromes in 1991.15 It also shows that the overall accident rate has reduced over the
period.
Between 1980 and 1991, there were 41 fatal, fire related, accidents. The majority of
those accidents occurred in areas remote from aerodromes served by an ARFFS. Fire
related accidents occurred near Bankstown, Perth and Essendon aerodromes and they
were served by an ARFFS.

Figure 4: Location of fire related fatal GA accidents, 1980 to 1991

5%
5%
5%

On or near a major aerodrome

On or near a GAAP aerodrome
On or near Other aerodrome
Remote from an aerodrome

85%

Between 1992 and 2003, there were 65 fatal, fire related, GA accidents. The majority
of those accidents occurred in areas remote from aerodromes/airports served by an
ARFFS but there was an increase in the number of accidents that occurred near
regional centre aerodromes and other localities. Fire related fatal accidents occurred at
Bankstown, Camden, Moorabbin and Jandakot general aviation aerodromes, plus near
Alice Springs, Ballarat, Bundaberg, Canberra, Ceduna, Coffs Harbour, Goulburn,
Groote Eylandt, Hamilton Island, Jerilderie, King Island, Mareeba, Melbourne, Mount
Gambier, North Stradbroke Island, Tamworth and Tindal aerodromes.

15
Poisson regression was used. Relative rate=1.48 for period 1992-2003 vs 1980-1989, p>.07, 95% CI
0.98 and 2.24.

9
Canberra, Alice Springs, Launceston and Tindal16 aerodromes were served by an
ARFFS.

Figure 5: Location of fire related fatal GA accidents, 1992 to 2003

5% 6%

23%
On or near a major aerodrome
On or near a GAAP aerodrome
On or near Other aerodrome
Remote from an aerodrome

66%

Between 1980 and 1991, there were 651 fatalities resulting from general aviation
accidents. Of those fatalities, 91 were from fire related accidents and were assessed to
be due to either fatal impact injuries or burns from the ensuing fire.

Figure 6: General aviation accident fatalities, 1980 to 1991

14%

Other fatalities
Fire related fatalities

86%

16
Tindal airport was served by an Australian Defence Force ARFFS.

10
Between 1992 and 2003, there were 542 fatalities resulting from general aviation
accidents. Of those fatalities, 120 were from fire related accidents and were assessed
to be due to either fatal impact injuries or burns from the ensuing fire.

Figure 7: General aviation accident fatalities, 1992 to 2003

22%

Other fatalities
Fire related fatalities

78%

There was a reduction in the number of fatalities in the 1992 to 2003 period compared
to the period from 1980 to 1991. But the number of fatalities in fire related accidents
increased as a proportion of the overall fatalities for the 1992 to 2003 period. There
was also a slight increase, during the 1992 to 2003 period, in the number of multiple
fire related fatalities compared with all fire related fatalities.17
The accidents involved either twin or single-engine aircraft with a maximum
allowable take-off weight less than 5,700 kg and that used aviation gasoline
(AVGAS) or propane as fuel.
Other than balloons, and some smaller aircraft, the majority of the aircraft were
constructed with a single aluminium outer layer on the fuselage that provides less
resistance to heat and flames than the majority of transport category aircraft used in
passenger operations. The post-impact fires were intense due to the availability of a
volatile fuel and the disruption of the aircraft structure during the impact sequence.
Information from recent accidents indicates that surviving occupants evacuated the
respective aircraft in less than 90 seconds of it coming to rest following ground
impact.

17
Changed from 43% of all fire related fatalities in 1980 to 1991 period to 51% in the 1992 to 2003
period.

11
Resistance of aircraft to burning
The US Federal Aviation Administration has conducted research and tests to establish
the resistance of transport category aircraft structures to the penetration of post-crash
fuel fires into the fuselage (burn-through). The burn-through resistance of the
aluminium skin of an aircraft is between 30 and 60 seconds. Tests using improved
thermal-acoustical material show significant improvements over insulation blankets
currently in use, which fail within 2 minutes. The tests also found that the means of
attaching thermal-acoustical material to the aircraft structure had a critical effect on
the effectiveness of the material.18 While work is under way to improve the
survivability during an accident, the focus is on transport category aircraft.19

ANALYSIS
The current ARFFS arrangements at Bankstown Airport do not comply with ICAO
SARPS, nor are they required to by Australian regulations or policy. Those
regulations and Government policy do not require an ARFFS to be provided for
operations at Bankstown Airport and consequently the NSW Brigades provide a
similar level of service to both the general community and the aerodrome.
Of the 94 aerodromes for which passenger data about scheduled air transport services
is gathered, 77 have a similar level of ARFFS to that provided to Bankstown Airport
and the other GAAP aerodromes across Australia.
In response to the accident notification, the first Brigades unit arrived about 9 minutes
after notification, which met its goal of providing assistance within 10 minutes of
notification, 90% of the time. While in the last 2 years it has only achieved 86% and
88% within 10 minutes, the average response time has improved by 1 minute (2002–
03 compared with 2001–02). The instructor pilot vacated the aircraft through the main
cabin door opening about 30 seconds after the aircraft came to rest. However, due to
the severity of his burn injuries, he died in hospital three and a half weeks later.
An earlier fire fighting response that met the ICAO standard, would in some
situations, enhance occupant survivability due to the availability of rescue and first
aid prior to the arrival of ambulance services. However, given the severity of the
injuries sustained by the instructor and student, and the fact that first aid was provided
very soon after the instructor vacated the aircraft, the provision of an ICAO standard
fire fighting service is unlikely to have improved the survivability of the accident.
On seeing that there was likely to be an accident, the ADC activated the wrong switch
due to the transposition of the location of the crash alarm switch in the Camden and
Bankstown tower consoles. The incorrect selection of the console power switch
resulted in a 30 second delay in the notification of the accident to emergency services.
That delay may have had some bearing on a response if an ARFFS had been located
at Bankstown Airport. However, under the circumstances the delay was considered to
be not significant in light of the overall time required for the Brigades’ units to reach
the accident site.

18
US Department of Transportation Federal Aviation Administration, Full-Scale Test Evaluation of
Aircraft Fuel Fire Burnthrough Resistance Improvements, January 1999.
19
C P Sarkos, Future Trends in Aircraft Fire Safety R&D, November 1998.

12
Similarly, the delay experienced by the first Brigade unit in gaining immediate entry
to the accident site within the aerodrome was not considered significant because of
the overall time taken for the unit to reach the site.
The ATSB Cross Modal Safety Comparisons discussion paper indicates that general
aviation is less safe than travel in transport category aircraft. There are many factors
that reduce the level of safety of persons involved in a general aviation aircraft
accident. While most relate to the severe forces involved in an aircraft crash, they also
include the lack of fire protection afforded by the aircraft structure, and the limited
window of opportunity for occupants to either vacate, or be rescued from, a burning
aircraft.
The implementation of recommendations from the House of Representatives Standing
Committee on Transport and Regional Services report Regional Aviation and Island
Transport Services: Making Ends Meet, particularly the formation of a working group
to advise on the strategic and optimal co-location of fire fighting services, and the
provision of fire fighting equipment and crew training to communities where fire
fighting services become co-located, should enhance occupant survivability in future
aviation accidents. However, due to the design limitations of general aviation aircraft
described above, any such service would need to achieve a significant reduction in the
current specified response times to ensure any improvement in occupant survivability.
While some enhancement in response times might be possible with the provision of a
location specific ARFFS (and commensurate FCC), there remains the problem with
the randomness of accident locations as indicated by the data on previous fire related
accidents. Since 1992, only 11% of fire related accidents have occurred on, or near
major or general aviation aerodromes. The majority of accidents (66%) occurred in
areas remote from an aerodrome with just under a quarter of fire related accidents
occurring on, or near regional and other aerodromes.
The results of research may eventually improve the burn-through resistance of aircraft
structures that will enhance transport category aircraft occupant survivability.
However, those improvements will take some time to flow through and provide
similar benefits to general aviation aircraft. In the interim, occupants of general
aviation aircraft might be able to improve survivability in a post-accident fire by
wearing clothing that offers greater protection from heat and flames.

CONCLUSION
The fire fighting arrangements for Bankstown Airport on 11 November 2003
complied with national policy. The response to the accident by the NSW Brigades
produced the expected result, in that resources to fight the fire were available within
the expected time. Despite the tragic loss of life, the arrangements were deemed to be
effective in terms of producing the intended or expected result for which they had
been established. The response to the accident was enhanced by the efforts of
personnel located on, and adjacent to the aerodrome.

13
SAFETY ACTION

Department of Transport and Regional Services safety action

The Department is considering whether an amended difference should be lodged with
International Civil Aviation Organization (ICAO) to ensure that any differences
between the national regulations for domestic aerodromes and ICAO standards and
recommended practices for aerodromes that conduct international operations are
apparent.

Airport Operator and New South Wales Fire Brigades’ safety action
Following a post-accident debrief with the Brigades, the Bankstown Airport operator
amended the airport emergency plan. In the future, if a safety officer is unavailable at
the assembly point when a Brigades’ unit arrives, that unit shall proceed unescorted
directly to an accident site within the aerodrome.

Airservices Australia safety action

Airport Services, a business unit of Airservices, has submitted a capital works
proposal to replace the Bankstown tower that will include the standardisation of crash
alarm and console power switches.

14