Bureau of Air Safety Investigation safety action
The Bureau is currently conducting an occurrence data analysis of accidents and incidents between 1989 and 1998 resulting from fuel exhaustion and/or fuel starvation. The results of this analysis are due for release in mid-1999.
CASA safety action
The March 1998 edition of CASA's quarterly journal "Flight Safety Australia", contained an article titled "Running on empty - too many pilots are running out of fuel". The article explored some of the underlying factors of fuel-related occurrences and provided advice to pilots on fuel planning and monitoring.
- The flight plan was poorly constructed and was not used to monitor the progress of the flight.
- The errors in the flight plan may have provided the pilot with a false impression that he had sufficient fuel for the flight.
- The pilot did not maintain an adequate sense of time, distance and fuel endurance.
- The engine failed due to fuel exhaustion.
- The pilot did not take advantage of suitable areas available for a forced landing.
- The aircraft stalled at a height from which recovery was not considered possible.
The flight plan used by the pilot for monitoring the flight was written in a manner that was not in accordance with conventional flight planning techniques. This meant that the recording of departure and arrival times, and the monitoring of the flight's progress, might not have been as straightforward as when using the conventional technique. No waypoints other than departure and destination points were entered, except for the significant sightseeing destinations of the Bungle Bungles and the East Alligator River area. The use of intermediate waypoints could have assisted the pilot in establishing arrival times and in monitoring the duration of the flights. A large error in the planned time interval between Kununurra and Tindal was considered to have been a significant factor in this accident and would have contributed to the pilot's overall lack of situational awareness during the accident flight. The overall plan allowed barely adequate time for refuelling at each landing point and made no allowance for time to be spent in sightseeing at significant locations. As a result, the day's activities were artificially shortened at the planning stage and the day's flying would inevitably take longer than originally planned.
Minimal notations of position and time on the flight documentation available suggested that the pilot was using the GPS as his main source of position and time reference. The balance of evidence also suggested that the pilot was not experienced in planning and conducting ad hoc scenic flights such as occurred on that day. It also suggested that the pilot did not monitor the relationship between the aircraft's position, time, and endurance remaining. There was no apparent recognition that the flight plan lacked the route segment between Timber Creek and Tindal.
The errors in the flight plan meant that the flight-planned fuel required for the Kununurra to Tindal flight, was inadequate. Considering that a more accurate time interval for the flight to Tindal would have been 3 hours 20 minutes, fuel should have been of considerable concern. The pilot apparently did not associate the delays in the flight and the additional leg from Timber Creek to Tindal with the amount of fuel available. On arrival in the Tindal circuit, the pilot did not appear to be aware that the fuel supply was almost exhausted, as none of the pilot's radio transmissions indicated any sense of urgency or need to land promptly.
When faced with the emergency, the pilot did not take advantage of available options that could have minimised the consequences. Once the engine power began to fail, it is likely that the pilot was concentrating on the reasons for the power loss and was attempting to restore power rather than considering suitable areas in which to make a forced landing. The pilot had not attempted to configure the aircraft for a forced landing as evidenced by the retracted landing gear and flaps at impact. The aircraft speed taken from the Tindal radar system recording and the witness reports of the aircraft "porpoising" prior to impact, are consistent with the aircraft stalling at approximately 300 ft AGL. With the aircraft in a stalled condition, the pilot had little control over the aircraft.
A Cessna 210 (C210), operating under visual flight rules, was
chartered for a one-day aerial sightseeing flight by a group of
four interstate visitors. The flight departed from Darwin NT and
flew to Kununurra WA where the aircraft was refuelled. It has been
established that the aircraft held fuel for approximately 240
minutes of flight when it departed Kununurra. This was consistent
with the fuel endurance noted on the flight plan.
From Kununurra, the aircraft flew to the Bungle Bungle Range
(Bungle Bungles) WA, where some scenic flying was carried out
before continuing to Timber Creek NT for an unscheduled landing due
to the unavailability of an air traffic control clearance into
Tindal airspace. During this leg of the flight the pilot amended
his SARTIME for arrival at Tindal by 1 hour 30 minutes. The
aircraft then flew to the Tindal airport at Katherine NT, where a
refuelling stop had been planned. The pilot again amended his
SARTIME for arrival at Tindal by a further hour. Approaching
Tindal, the pilot communicated with other aircraft in the area and
manoeuvred to establish a traffic pattern. After further
communications, the pilot advised that he was joining downwind for
runway 14. Shortly after this, the aircraft was observed to be
flying at a very low height with the engine "spluttering".
Witnesses saw the aircraft "porpoising" as it descended into trees.
The sound of an impact was heard shortly after. The aircraft speed
taken from the Tindal radar system recording was consistent with
the aircraft being in a stalled condition from approximately 300 ft
AGL.
The accident site was located approximately 6.6 km west of
Tindal airport. Adjacent to the accident site were a number of
areas suitable for a forced landing. Apart from these areas, the
terrain was generally flat with occasional rocky outcrops, and was
moderately treed. The accident site was contained within the
moderately treed area. The five occupants had received fatal
injuries.
Pilot information
The pilot held a commercial flight crew licence, a valid class
one medical certificate and appropriate endorsements to allow him
to operate C210 aircraft. The operator employed the pilot on a
casual basis.
At the time of the accident, the pilot had accrued a total of
798 hours flying experience. Of this, 636.8 hours were in command
with 481.3 hours on C210 aircraft. In the previous 30 days he had
flown 47.2 hours, all of which were conducted as pilot in command
on the C210. Evidence indicated that the last time the pilot had
demonstrated practice forced landings was in March 1996. Evidence
also indicated that the pilot had not previously conducted an
extended flight with fare-paying passengers as was planned on the
day of the accident.
Wreckage examination
The combination of speed and descent angle of the aircraft
resulted in an estimated descent rate at impact of about 3,100
ft/min and an estimated peak impact load of 64G. Such an impact was
not considered to be survivable.
The right fuel tank had been ruptured at its inboard end and
there was evidence that a small amount of fuel had been released
from the ruptured area. The left fuel tank was intact and it
contained about one cup of fuel. Traces of fuel were found in some
areas of the engine fuel system, most areas being dry. The aircraft
was assessed as containing no useable fuel at the time of impact.
The engine instruments indicated that the aircraft had flown for
approximately 240 minutes since refuelling at Kununurra.
The landing gear and flaps were retracted and there was no
evidence that either had been selected in anticipation of
configuring the aircraft for a landing.
Flight Planning
On the reverse side of the flight plan form, the pilot had
documented the planned flight legs for each route. Each line
commenced with the departure point and contained the data for the
flight to the destination. This method was not in accordance with
the conventional method of flight planning and meant that the
thought process required for each line of data was to think FROM
the location at the beginning of the line rather than TO the
location at the beginning of the line.
The line on the flight plan commencing with "KU" contained data
applicable for the flight from Kununurra to the Bungle Bungles, a
distance of 100 NM and an estimated elapsed time of 45 minutes. No
time had been planned for sightseeing at the Bungle Bungles.The
line on the flight plan commencing with "BU", assumed to be the
Bungle Bungles, contained data applicable for the flight from the
Bungle Bungles to Timber Creek, a distance of 161 NM and an
estimated elapsed time of 1 hour 12 minutes. No data had been
entered on the flight plan for the leg between Timber Creek and
Tindal, a distance of 131 NM, which would have been expected to
take approximately 60 minutes. A more accurate planned time
interval, calculated by the investigation team, for the flight
between Kununurra and Tindall was approximately 3 hours 20 minutes,
including an allowance of 20 minutes for flight in accordance with
standard operating procedures for sightseeing at the Bungle
Bungles. The pilot's flight plan indicated that he had allowed only
1 hour and 57 minutes. No waypoints other than departure and
destination points were entered, except for the significant
sightseeing destinations of the Bungle Bungles and the East
Alligator River area. The anomaly in the flight-planned time
between Kununurra and Tindal was consistent with the information
contained on the other side of the flight plan form and which was
submitted to Airservices Australia for flight notification
purposes. Although the actual time of landing at Timber Creek could
not be established, evidence suggested that the actual time from
Kununurra to Timber Creek was probably about 2 hours 50 minutes.
Although two fuel calculation columns were annotated with "KU" and
"TN" under the "endurance" line at the bottom of the table, no
other entries were made in the columns.
Fuel considerations
Persons interviewed during the investigation said that the pilot
considered fuel loading prior to flights and there had been
occasions when the aircraft's load was reduced in order to carry an
adequate fuel load. However, there was some evidence, other than on
the accident flight, that the pilot was not in the habit of
maintaining a running log of time, distance and fuel endurance. The
operator's chief pilot advised that during the pilot's most recent
check, he noted that the pilot used only the GPS for navigation and
did not keep a flight log, but that he always seemed to know where
he was.
There was no evidence to suggest that the fuel gauges were
unreliable. A video recorder recovered from the wreckage contained
a recording of the flight taken by one of the passengers and
provided a clear view of the instrument panel while the aircraft
was on approach to land at Timber Creek. The left fuel gauge
indicated that the tank was nearly empty while the right gauge
indicated approximately one quarter full. Calculations based on
this evidence determined that the aircraft had approximately 15 US
gallons or some 57 L on board during the approach to Timber Creek,
which equated to approximately 1 hour's flight time.
The operator
The chief pilot was the owner of the organisation and the only
full-time employee. Three part-time pilots, including the pilot of
the accident flight, were employed on a casual basis. From June
1994, the operator's AOC was upgraded from aerial work to include
charter operations.
The company operations manual required the chief pilot to ensure
that pilots rostered for charter flights familiarise themselves
with the route and conditions expected along the route. The chief
pilot reported that he had not seen the pilot's plan, and that he
had not discussed the plan or the route in detail with the pilot
before he departed on the accident flight.
Navigation
One Global Positioning System (GPS) unit was recovered from the
wreckage and another was found in the pilot's bag. The unit in use
on the accident flight contained waypoints appropriate to the
flights conducted on the day of the accident; however, the
route-tracking facility was not selected. Although the unit could
be used for fuel endurance calculations, no parameters had been
entered into the unit to allow this to occur. The last position
retrieved from the unit after the accident was consistent with the
accident location.
Information obtained during the investigation suggested that the
pilot tended to use the GPS exclusively for navigation, and was not
in the habit of recording any fixes on a map or other
documentation. This information was consistent with the evidence
provided by the pilot's documentation recovered from the aircraft.
No entries had been made on the pilot's navigation charts. Only the
intended tracks had been drawn on them. Some positions and
estimates given by radio would most likely have been obtained by
reference to a GPS unit.
The flight plan contained almost no information other than that
entered during the flight planning stage. No positions and times
were entered. There was no evidence of any attempt to maintain a
fuel log by recording fuel quantities at significant points, or
recording tank selection times.