Sequence of events

On 7 May 2005, a Fairchild Aircraft Inc. SA227-DC Metro 23 aircraft, registered VH•

TFU, with two pilots and 13 passengers, was being operated by Transair on an

instrument flight rules (IFR) regular public transport (RPT) service from Bamaga to

Cairns, with an intermediate stop at Lockhart River, Queensland. At 1143:39 Eastern

Standard Time, the aircraft impacted terrain in the Iron Range National Park on the

north-western slope of South Pap, a heavily timbered ridge, approximately 11 km north•

west of the Lockhart River aerodrome.

At the time of the accident, the crew was conducting an area navigation global

navigation satellite system (RNAV (GNSS)) non-precision approach to runway 12. The

aircraft was destroyed by the impact forces and an intense, fuel-fed, post-impact fire.

There were no survivors.

The accident was almost certainly the result of controlled flight into terrain; that is, an

airworthy aircraft under the control of the flight crew was flown unintentionally into

terrain, probably with no prior awareness by the crew of the aircraft’s proximity to

terrain.

Weather conditions in the Lockhart River area were poor and necessitated the conduct

of an instrument approach procedure for an intended landing at the aerodrome. The

cloud base was probably between 500 ft and 1,000 ft above mean sea level and the

terrain to the west of the aerodrome, beneath the runway 12 RNAV (GNSS) approach,

was probably obscured by cloud.

The flight data recorder (FDR) data showed that, during the entire descent and

approach, the aircraft engine and flight control system parameters were normal and that

the crew were accurately navigating the aircraft along the instrument approach track.

The FDR data and wreckage examination showed that the aircraft was configured for

the approach, with the landing gear down and flaps extended to the half position. There

were no radio broadcasts made by the crew on the air traffic services frequencies or the

Lockhart River common traffic advisory frequency indicating that there was a problem

with the aircraft or crew.

Crew performance

As the copilot was making the radio broadcasts during the approach, it is very likely that

the 40-year old pilot in command was the handling pilot. The pilot in command was

Transair’s base manager at Cairns and an experienced Metro pilot. However, given the

relatively complex type of approach being flown, he would have been reliant on the

relatively inexperienced 21-year old copilot to assist with the high cockpit workload.

There was a significant potential for crew resource management problems within the

crew in high workload situations, given that there was a high trans-cockpit authority

gradient and neither pilot had previously demonstrated a high level of crew resource

management skills.

The crew commenced the Lockhart River Runway 12 RNAV (GNSS) approach, even

though the crew were aware that the copilot did not have the appropriate endorsement

and had limited experience to conduct this type of instrument approach. A nondirectional

beacon approach was also available at Lockhart River, and both pilots were

endorsed for that approach. Despite the weather and copilot inexperience, the pilot in

– xiii –

command used descent and approach speeds and a rate of descent greater than specified

for the aircraft in the Transair Operations Manual, and exceeding those appropriate for

establishing a stabilised approach.

During the approach, the aircraft descended below the segment minimum safe altitude

for the aircraft’s position on the approach. The aircraft’s high rate of descent, and the

descent below the segment minimum safe altitude, were not detected and/or corrected

by the crew before the aircraft collided with terrain.

While the investigation was complicated by an inoperative cockpit voice recorder, no

witnesses, and the extent of destruction of the aircraft, it determined that the crew

probably experienced a very high workload during the approach and probably lost

situational awareness about the aircraft’s position along the approach path.

The pilots’ aircraft endorsements, clearance to line operations, and route checks did not

meet all the relevant regulatory and operations manual requirements to conduct RPT

flights on the Metro aircraft. However, these limitations were not considered to have

had an influence on the conduct of the flight.

Ground proximity warning system

There was no evidence that the ground proximity warning system (GPWS) was not

functioning as designed. Simulation by the GPWS manufacturer indicated that the crew

should have received a one second ‘terrain terrain’ alert about 25 seconds prior to

impact, followed by a second ‘terrain terrain’ alert and a continuous ‘pull up’ warning

for the final 5 seconds of flight. However, research has shown that the alerts and

warnings in the final 5 seconds of flight would not have been sufficient for the crew and

aircraft to effectively respond to the GPWS annunciations.

A terrain awareness and warning system (TAWS, commonly referred to as enhanced

GPWS) provided advantages over standard GPWS. It enhanced pilot situational

awareness by providing coloured terrain information on a continuous terrain display in

the cockpit and providing more timely alerts and warnings. Had the aircraft been fitted

with a TAWS, it is probable that the accident would not have occurred.

Transair processes

In addition to the substantive crew actions and local conditions that contributed to the

accident, the investigation identified a number of safety factors relating to Transair that

contributed to the accident. In particular, the flight crew training program had

significant limitations, such as superficial or incomplete ground-based instruction

during endorsement training, no formal training for new pilots in the operational use of

global positioning system (GPS) equipment, no structured training on minimising the

risk of controlled flight into terrain, and no structured training in crew resource

management (or human factors management) and operating effectively in a multi-crew

environment.

Transair’s processes for supervising the standard of flight operations at the Cairns base

had significant limitations, such as not using an independent approved check pilot to

review operations, reliance on passive measures to detect problems, and no defined

processes for selecting and monitoring the performance of the base manager. In

addition, Transair’s standard operating procedures for conducting instrument

approaches had significant limitations, such as not providing clear guidance on

approach speeds, not providing guidance for when to select aircraft configuration

changes during an approach, no clear criteria for a stabilised approach, and no

– xiv –

standardised phraseology for challenging safety-critical decisions and actions by other

crew members.

Transair’s organisational structure, and the limited responsibilities given to nonmanagement

personnel, resulted in high work demands on the Transair chief pilot. This

resulted in a lack of independent evaluation of training and checking, and created

disincentives and restricted opportunities within Transair to report safety concerns with

management decision making. There was no structured process within Transair for

proactively managing safety-related risk associated with its flight operations.

Furthermore, the chief pilot did not demonstrate a high level of commitment to safety

and appeared to be over-committed, with additional roles as chief executive

officer/managing director of the company, the primary check and training pilot, and

working regularly in Papua New Guinea for an associated company.

In addition, limitations were also identified with Transair’s flight crew proficiency

checking program and the useability of the Transair Operations Manual. However,

these issues were not considered to be contributing safety factors to the accident.

Regulatory oversight

The investigation also identified contributing safety factors relating to the regulatory

oversight of Transair by the Civil Aviation Safety Authority (CASA). In particular,

CASA did not provide sufficient guidance to its inspectors to enable them to effectively

and consistently evaluate several key aspects of operators' management systems. These

aspects included evaluating organisational structure and staff resources, evaluating the

suitability of key personnel, evaluating organisational change, and evaluating risk

management processes. CASA also did not require operators to conduct structured

and/or comprehensive risk assessments, or conduct such assessments itself, when

evaluating applications for the initial issue or subsequent variation of an Air Operator’s

Certificate.

In addition, CASA’s oversight of Transair, in relation to the approval of Air Operator’s

Certificate variations and the conduct of surveillance, was sometimes inconsistent with

CASA’s policies, procedures and guidelines. However, this was not considered to have

been a contributing safety factor.

Other safety factors

The investigation also identified a range of other safety factors which did not meet the

definition of a contributing safety factor or which could not be as clearly linked to the

accident because of lack of evidence, but which were still considered to be important to

communicate in an investigation report with a focus on future safety. In addition to

some aspects of Transair’s processes and regulatory oversight activities, these safety

factors related, among other things, to the possibility of poor intra-cockpit

communication, instrument approach design, instrument approach chart presentation,

and regulatory requirements.

The Australian convention for waypoint names in RNAV (GNSS) approaches did not

maximise the ability to discriminate between waypoint names on the aircraft GPS

display and/or on the instrument approach chart. In addition, there were several design

aspects of the Jeppesen RNAV (GNSS) approach charts, which were very likely to have

been used by the crew, that could lead to pilot confusion or a reduction in situational

awareness. These included limited reference regarding the ‘distance to run’ to the

missed approach point, mismatches in the vertical alignment of the plan-view and

profile-view on charts such as that for the Lockhart River runway 12 approach, use of

– xv –

the same font size and type for waypoint names and altitude limiting steps, and not

depicting the offset in degrees between the final approach track and the runway

centreline. There were also limitations in the terrain information provided on Jeppesen

instrument approach charts.

CASA’s process for accepting an instrument approach did not involve a systematic risk

assessment of pilot workload and other potential hazards and warnings, including

activation of a GPWS. There was also no regulatory requirement for instrument

approach charts (including the Lockhart River Runway 12 RNAV (GNSS) approach

chart) to include coloured contours to depict terrain as required by the International

Civil Aviation Organization (ICAO) Annex 4, to which Australia had not notified a

difference.

Although CASA released discussion papers in 2000, and further development had

occurred since then, there was no regulatory requirement for initial or recurrent crew

resource management (CRM) training or for RPT operators to have a safety

management system. In addition, there was no regulatory requirement for flight crew

undergoing a type rating on a multi-crew aircraft to be trained in procedures for crew

incapacitation and crew coordination including allocation of pilot tasks, crew

cooperation and use of checklists. This was required by ICAO Annex 1, to which

Australia had notified a difference.

The investigation also determined that CASA’s guidance material provided to operators

about the structure and content of an operations manual was not as comprehensive as

that provided by ICAO in areas such as multi-crew procedures and stabilised approach

criteria, and that its process for evaluating the content of an operations manual did not

consider the useability of the manual, particularly in electronic format. There was also

no regulatory requirement for multi-crew RPT aircraft to be fitted with a serviceable

autopilot.

Safety action

This investigation identified important learning opportunities for pilots, operators and

regulatory agencies to improve future aviation safety and to seek to ensure such an

accident never happens again. During the course of the investigation, the ATSB issued

10 safety recommendations and encouraged other safety action.

Safety action has been taken by several organisations to address the safety issues

identified during this investigation. A number of additional safety recommendations

were issued by the ATSB, including seven recommendations to the Civil Aviation

Safety Authority on its regulatory oversight activities and regulatory requirements.

Recommendations on aspects of instrument approach charts were also issued to

Airservices Australia and Jeppesen Sanderson Inc.

The ATSB did not issue recommendations regarding the serious safety issues of the

operator because Transair had surrendered its Air Operator’s Certificate on 4 December

2006 and ceased to operate.

– xvi –