Conference PaperPDF Available

Review of Transport Aircraft Ditching Accidents

Authors:

Abstract and Figures

Aircraft ditching, i.e. the controlled emergency landing on water, is of relevance for the design and certification of transport aircraft. Ditching investigations are generally based on scaled model tests as well as numerical simulations. Supplementary, the analysis of ditching accidents is another valuable investigation method. Although water impact accidents of aircraft are rare and cases of passenger transport jet aircraft ditching are almost none existent, some documented examples exist, which provide insight into the impact scenario. The paper presents a compilation of all known planned ditching events of passenger transport jet aircraft based on available accident data and reports. Emphasis is given to two recent planned ditching events of a Boeing 767 (in 1996) and a Boeing 737 (in 2002), which are analyzed in detail from an engineering point of view. As a typical example of an unplanned ditching, the accident of a Boeing 707 (in 2000) is investigated. The focus of the three case studies is put on extracting the impact conditions in view of the related hydrodynamic loads and analyzing the respective damages to the aircraft structure. Based on the investigated scenarios, the paper concludes the key factors for a successful aircraft ditching as well as those factors contradicting it and outlines remaining uncertainties. Moreover, the examples included highlight the crucial aspects of modeling and simulating aircraft ditching. The conducted accident analysis thus enhances the knowledge about transport aircraft ditching and provides guidance to goaloriented ditching investigations.
Content may be subject to copyright.
Review of Transport Aircraft Ditching Accidents
Olaf Lindenau (lindenau@tu-harburg.de), Thomas Rung (thomas.rung@tu-harburg.de)
Institute for Fluid Dynamics and Ship Theory (M8), Hamburg University of Technology
Abstract
Aircraft ditching, i.e. the controlled emergency landing on water, is of relevance for the design and certi-
fication of transport aircraft. Ditching investigations are generally based on scaled model tests as well as
numerical simulations. Supplementary, the analysis of ditching accidents is another valuable investigation
method. Although water impact accidents of aircraft are rare and cases of passenger transport jet aircraft
ditching are almost none existent, some documented examples exist, which provide insight into the impact
scenario.
The paper presents a compilation of all known planned ditching events of passenger transport jet aircraft
based on available accident data and reports. Emphasis is given to two recent planned ditching events
of a Boeing 767 (in 1996) and a Boeing 737 (in 2002), which are analyzed in detail from an engineering
point of view. As a typical example of an unplanned ditching, the accident of a Boeing 707 (in 2000) is
investigated. The focus of the three case studies is put on extracting the impact conditions in view of the
related hydrodynamic loads and analyzing the respective damages to the aircraft structure.
Based on the investigated scenarios, the paper concludes the key factors for a successful aircraft ditching
as well as those factors contradicting it and outlines remaining uncertainties. Moreover, the examples
included highlight the crucial aspects of modeling and simulating aircraft ditching. The conducted accident
analysis thus enhances the knowledge about transport aircraft ditching and provides guidance to goal-
oriented ditching investigations.
Keywords: aircraft ditching, unplanned ditching, water impact, impact damage, impact loads, accident
investigation
Contents
1 Introduction 2
2 Evaluation of transport jet aircraft ditching events 4
3 Ditching accident case studies 5
3.1 Planned ditching of Boeing 767 in 1996 off Comoros Islands . . . . . . . . . . . . . . . . . . . . . . 5
3.1.1 History of flight and analysis of water impact . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1.2 Summary from ditching point of view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Unplanned ditching of Boeing 707 in 2000 in Lake Victoria . . . . . . . . . . . . . . . . . . . . . . . 8
3.2.1 Historyofight .......................................... 8
3.2.2 Analysis of flight data recorder parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2.3 Damages to aircraft structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.4 Analysis of accident from ditching point of view . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2.5 Comparison of ditching tests for this aircraft type to the unplanned ditching accident . . . . . 12
3.3 Planned ditching of Boeing 737 in 2002 in Bengawan Solo River . . . . . . . . . . . . . . . . . . . . 13
3.3.1 Historyofight .......................................... 13
3.3.2 Analysis of damages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3.3 Evacuation and injuries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.3.4 Analysis of accident from ditching point of view . . . . . . . . . . . . . . . . . . . . . . . . . 16
4 Assessment of planned and unplanned ditching events 17
5 Conclusion 18
Bibliography 19
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 1/19
1 Introduction
Water impact accidents of aircraft can be divided into two groups: ditching accidents of the airborne aircraft and
overrun accidents with prior ground contact during take-off or landing. Moreover, the definition excludes all non
survivable accidents where the aircraft also impacts a body of water but without a chance of survival. Ditching
is further classified in planned ditching and unplanned ditching. The general aviation notion of “ditching”
refers to an emergency landing on water as a planned water contact. The crew, with the aircraft under control,
deliberately lands on water according to recommended procedures given in the flight manual. The destruction
of the aircraft is accepted, provided the crew and passengers can safely escape and be rescued. As opposed to
this, unplanned ditchings refer to water impacts without appropriate preparation. Consequently, only some, if
any at all, measures of the recommended ditching procedure can be followed.
For ditching performance analysis of aircraft a ditching event can be subdivided into four major consecutive
phases (Fig. 1): approach, impact, landing and floatation phase.
Approach phase Impact phase Landing phase Floatation phase
approach conditions
environmental
conditions
hydrodynamic
impact pressures
local deformations
motion after impact
planing condition
global dynamic
response of structure
detachment of
components
floatation position
water ingress
evacuation
Figure 1: Phases of aircraft ditching.
The approach phase covers the approach conditions for the airborne aircraft close to the water surface, which
are then the initial conditions for the impact scenario. The approach conditions consist of the aircraft’s state
of motion such as flight path angle, pitch attitude and aircraft speed. Moreover, the aircraft’s mass properties
regarding weight, center of gravity and inertia as well as the flap setting, landing gear position, and engine
thrust are of relevance. The environmental conditions contain wind and sea state. The decisive aspect about
of the approach phase is to reduce the energy which has to be absorbed and transformed during the impact
phase to a minimum. The reduction of the kinetic energy at the end of the approach phase can be achieved by
reducing the aircraft weight by possible fuel jettisoning and, if possible, by using a maximum-flap configuration
to lower the speed. Above all, the forward speed should be minimized while keeping a low to moderate sink
rate, as the velocity directly determines the hydrodynamic pressure.
The impact phase yields local deformations along the fuselage due to the hydrodynamic pressure. The landing
phase covers the motion of the aircraft including the accelerations acting on the passengers and structural
components of the aircraft. Critical situation occur when the aircraft nose dives into the water or the aircraft
skips on the water surface. Accordingly, loads and accelerations exceed tolerable and survivable thresholds.
Aspects of interest are the global dynamic impact response of the structure, as well as the possible detachment
of aircraft components such as engines and flaps. Once the airplane comes to rest, the floatation phase starts,
which is governed by water ingress and likely ends with the sinking of the airplane. The evacuation of the
passengers and crew of the aircraft takes place within the floatation phase.
The ditching of aircraft can be investigated by ditching accident analysis, by ditching tests and numerical
simulations. Each of these three methods individually contributes to the assessment of ditching of aircraft and
also the methods themselves are linked to each other. The approach is depicted in Fig. 2 as the “triangle of
aircraft ditching investigation methods”. Ditching accident analysis refers to the technical investigation of actual
transport aircraft water-impact accidents comprising their causes, circumstances, and consequences. Physical
ditching tests are possible using prepared full-scale aircraft or scaled models. Model-scale ditching tests are the
traditional way of ditching performance analysis of aircraft designs. The level of complexity of numerical ditching
simulations ranges from simple analytical water-impact formulas to numerical simulation methods taking into
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 2/19
Analysis
Accident
Ditching
Ditching
Simulation of
Numerical
Ditching
Ditching
of
Aircraft
Tests
Figure 2: The triangle of aircraft ditching investigation methods.
account the complete aircraft and the near field of the surrounding fluids. While the ditching accident analysis
is confined to a few aircraft under special conditions, ditching tests and numerical simulations of ditching allow
to investigate aircraft ditching performance during the design phase considering various parameter influences.
While generally ditching accident analysis within the official accident investigation is aimed at determining the
causes of the accident, this paper is devoted to the ditching itself. As the key issue in ditching of aircraft is
the impact on water, this paper focuses on the hydrodynamics of ditching, but also includes aerodynamic and
mechanic aspects.
There are a number of constraints in the analysis of aircraft ditching accidents:
As Table 1 shows ditchings are very rare events and occur for different aircraft under different circum-
stances. Thus, no meaningful statistical evaluation is possible. Alternatively, the ditching accident analysis
has to be conducted for individual accidents or by clusters of selected cases.
In hardly any ditching accident the flight data recorder data of the water impact are available. Besides
that, reliable observations of the first ditching phases and their documentation are rare. Hence, the aircraft
state during approach and the sequence of events during impact and landing needs to be estimated. This
estimation of leads to uncertainties regarding the actual conditions in the first three ditching phases.
Usually accident investigators are examining the final or post-final situation at the crash site. Next to the
documentation of the state of the aircraft after the crash, they also force the difficulty to determine the
state of the aircraft before the crash. In some cases videos and photos of the accident itself or at least of
some of the four ditching phases are available. In addition to measured weather data the visual material
reveals the environmental conditions.
Not necessarily all material gathered during the official accident investigation is released to the public.
Mind, that the focus of the accident investigation is usually put on the cause of the accident. Apart from
this, it has to be noted that the accident investigation and the published report reflects a certain view
of a group of people on an event at a point in time. This view is affected by scientific (e.g. analytical,
technological, and recovery capabilities) as well as non-scientific factors (e.g. significance of event, ex-
perience of investigators, political climate, relationship with regulatory authorities). The present study
- like others - is condemned to rely on spares publicly available data. This basically consists of the official
accident report and additional visual material.
Unlike numerical simulations and ditching tests, the accident analysis offers no direct insight into the
physics of ditching and thus makes it difficult to reconstruct a cause-and-effect chain. As these kind of
data are only available for existing aircraft they can only be used for future aircraft within the design
stage.
Despite the above mentioned constraints the analysis of ditching accidents gives irrefutable evidence of what
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 3/19
happens in aircraft ditching. Thus, it shows what should be considered in numerical and experimental ditching
investigations in the design phase of an aircraft. Accident analysis also reveals how regulations as well as
experimental and numerical investigation methods match the actual accident scenarios. Ditching tests and
numerical simulations of ditching may serve to investigate the relation between the impact conditions and the
outcome of an actual accident in terms of motions, loads, and damages. Thus, ditching tests and numerical
simulations can enhance the understanding and facilitate the analysis of such accidents.
2 Evaluation of transport jet aircraft ditching events
Table 1 summarizes the short list of passenger transport jet aircraft ditching accidents. All listed events accidents
classify as ditchings due to the fact that the flight crew was aware of the necessity to ditch the aircraft and there
was some time to prepare the passengers and the aircraft. The list explicitly excludes propeller driven as well
as military and general aviation aircraft ditching events in order to focus on the ditchings of large passenger
transport jet aircraft.
Table 1: Planned ditching events of passenger transport jet aircraft.
Date
A/C Type Location On board /
Fatalities Description
21 Aug. 1963
TU-124
Neva River, St. Peters-
burg, Russian Federation
52 / 0 Fuel exhaustion on diverted flight with
non retracted undercarriage.
2 May 1970
DC-9
Carribean Sea, NE of St.
Croix, Virgin Islands (US)
63 / 23 Fuel exhaustion resulting from
continued, unsuccessful landing attempts.
11 Sep. 1990
B-727
Atlantic Ocean, SE off
Newfoundland, Canada
18 / 18 Contact lost following fuel problem
distress call; wreckage not located.
23 Nov. 1996
B-767
Indian Ocean, off Grande
Comore, Comoros
175 / 125 Fuel exhaustion due to interference by
hijackers.
16 Jan. 2002
B-737
Bengawan Solo River,
near Klaten on Java
Island, Indonesia
60 / 1 Engine failure in thunderstorm and
unsuccessful restart attempts.
15 Jan. 2009
A320
Hudson River, New York,
United States
155 / 0 Dual engine failure caused by bird strike.
The first event refers to the Tupolev TU-124 ditching in 1963. The accident has been reported by Flight
International (1964) and data has been compiled and discussed by Townshend (1965). A number of favorable
aspects lead to the successful outcome of the ditching: the pilot handling of the ditching, the environmental
conditions regarding weather and visibility, and the immediate rescue by a tug boat towing the aircraft to shore.
Furthermore, the light weight of the aircraft due to empty fuel tanks, as well as the little damage experienced by
the aircraft structure at water impact which preserved the floatation capability of the aircraft, have contributed
to the positive outcome.
The McDonnell Douglas DC-9 ditching in 1970 was investigated by the National Transportation Safety Board
(NTSB) (1971). Additionally, an extensive analysis of this accident was conducted by the NTSB (1972).
Despite the great search and rescue efforts conducted and reported by the Joint Rescue Coordination Centre
(RCC) Halifax (1990) in connection with the Boeing B-727 ditching in 1990, no parts of the aircraft could be
located with reasonable certainty. The flight crew had made a distress call to two other aircraft stating that
they were low on fuel and planned to ditch the aircraft on the ocean (Fink, 1990).
Due to the availability and wide distribution of a video recording, the Boeing B-767 ditching in 1996 is well
known. The ditching will be discussed as case study in section 3.1.
The ditching of the Boeing B-737 in 2002 is the second most recent ditching of a jet transport aircraft. This
case proofs that modern type aircraft can be successfully ditched and will be discussed in section 3.3.
The ditching of the Airbus A320 in 2009 was at the time of writing under investigation by the National
Transportation Safety Board (NTSB) (2009).
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 4/19
Along with the development of modern transport aircraft a number of studies regarding water-impact accidents
have been performed. These compile the data of a number of accidents in order to identify technical and
procedural deficiencies. The studies aim to give recommendations for improvements and provide guidance
to adapt regulations. The latest of such studies are a review of transport aircrew training programs related
to ditching by Cosper and McLean (1998) and a two-part study on “Transport water impact and ditching
performance” sponsored by the Federal Aviation Administration (FAA) in 1996. The motivation were the
transport aircraft water impact accidents occurring in the period 1959 to 1994, and the rising number of
overwater operations of transport aircraft and the associated potential risk for water-impact accidents. The
first part of the FAA study was performed by Patel and Greenwood (1996) having three objectives: first to
review and analyze worldwide transport accident data relative to water impacts and ditching performance,
second to compare the results of the study with current FAA requirements to determine their adequacy and
relevancy, and third to conduct a survey of major worldwide airports to determine their proximity to water.
The second part of the FAA study was conducted by Tahliani and Muller (1996) and comprises three topics:
accident data analysis, airport water rescue and emergency floatation equipment.
The ditching accidents of the B-767 in 1996 and of the B-737 in 2002 are not included in the previously published
accident studies. Thus, the present paper reviews these accidents. Additionally, the unplanned ditching of a
Boeing B-707 ditching in 2000 is analyzed in section 3.2 in order to widen the basis for the present study and
as a typical scenario for an unplanned ditching, i.e. descent into a body of water short of the runway during
approach.
3 Ditching accident case studies
3.1 Planned ditching of Boeing 767 in 1996 off Comoros Islands
On 23 Nov. 1996 at 12.20 UTC (15.20 local time) the Boeing 767-200ER Ethiopian Airlines Flight No. 961
with registration ET-AIZ ditched off the island Grande Comore, Union of the Comoros, in the Indian Ocean.
The information presented here is extracted from the aircraft accident report by the Ethiopian Civil Aviation
Authority - Flight Safety Department (1998). The hijacking situation has been analyzed by the Federal Aviation
Administration (FAA) (1996). The accident is widely known because of the extensive news coverage following
the accident, which includes a short video taken by a tourist from the beach at the crash site of the water
impact and displays the disintegration of the aircraft (e.g. CNN interactive, 1996a,b,c).
The Boeing 767-200ER is a mid-size, wide-body twin-jet airliner in an extended-range variant. The fuselage
with a length of 48.5m has a height of 5.41m and a width of 5.03m. The B-767 features a low-wing cantilever
monoplane of 47.6 m span and a conventional tail unit with a single fin and rudder. The aircraft is powered by
two under wing mounted turbofan engines. The B-767-200ER with registration ET-AIZ was configured with
190 seats in twin-aisle arrangement: 12 first-class, 24 business-class, and 154 economy-class seats.
3.1.1 History of flight and analysis of water impact
The Boeing 767 with 163 passengers and 12 crew members was hijacked en-route from Addis Ababa, Ethiopia to
Abidjan, Ivory Coast. The hijackers demanded to be flown to Australia and ignored the necessity of refueling.
The aircraft eventually ran out of fuel and the pilot ditched the aircraft 500m away from the shore in the
northern part of Grande Comore. The aircraft broke into four parts. 119 passengers and 6 crew members were
fatally injured. As cause of the accident “unlawful interference by hijackers which resulted in loss of engines
power due to fuel exhaustion” is given in the accident report.
Due to the total engine power loss the digital flight data recorder (DFDR) and the cockpit voice recorder (CVR)
stopped recording about 25 minutes before the impact. Thus, no data of the ditching can be concluded from
the DFDR. The final approach and the ditching of the aircraft was video recorded by a tourist (Fig. 3).
According to the accident report an altimeter reading of 150ft and an airspeed of 200kts were obtained less
than 2 minutes before impact. Pilot and first officer had been left alone by the hijackers to assume control by
this time. Before that the hijackers had partly made improper control inputs. Further, the report reveals that
the flight crew had the intention to turn the aircraft to the left in order to ditch parallel to the wave crests.
Regarding the impact motion of the aircraft, the report states that “the aircraft brushed the water in a left-wing
low attitude and then was held in a straight and level attitude. After hitting a reef with its belly, it broke into
four sections.” The fuselage broke behind the cockpit section, in front of the wing box and near the middle of
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 5/19
a) b) c)
Figure 3: Sequence of still images taken from tourist video of Boeing 767 ET-AIZ ditching in 1996 off Grande
Comore: a) first water impact of the left wing, b) almost straight and leveled aircraft during the landing phase,
c) disintegration due to impact with a reef.
the aft fuselage part, Fig. 4. The left engine was found still attached to the wing and the right engine separated
from the wing. The center section came to rest upside down. The impact to the reef in shallow water is given
as reason for the destruction of the aircraft.
Figure 4: Boeing 767 ET-AIZ ditching in 1996 off Grande Comore: fuselage breakage (bottom) and seating
arrangement with coloring according to injuries (top). Graphics taken from accident report by the Ethiopian
Civil Aviation Authority - Flight Safety Department (1998) and colored according to injuries.
With respect to the configuration of the aircraft at impact, the accident report states, that “the landing gears
were in a retracted position with the doors closed. The RAM air turbine was deployed. The wing flaps were
up.” The weight of the aircraft at impact was the actual take-off weight of 124.419t reduced by the consumed
fuel. The weather at the accident site at the time of the accident is estimated with NE wind of 3 Bft, 30C air
and in the order of 25C water temperature.
The RAM air turbine provided power only for the stand-by-instruments. Thus “there was no vertical speed
information required to maintain a 200-300fpm rate of descent, in accordance with the ditching procedure.”
Instructions on ditching were provided by the Flight Operation and Flight Crew Training manual. In case of
ditching the crew is expected to complete the ditching preparation checklist in the Quick Reference Handbook.
But the circumstances of this accident did not allow the crew to perform in accordance with the procedure. “In
addition, the procedures do not provide for ditching with both engines out.”
There was an early awareness of the possibility of having to ditch among the flight crew, the area control and
the hijackers. The pilot already discussed with the area control and hijackers the need for refueling and the
otherwise possible ditching about three hours before the accident. About half an hour before the impact the pilot
made an announcement speaking of an expected “crash landing”. No additional pre-emergency briefing of the
passengers was carried out. Only the announcement to sit down and fasten the seat belts could be given by the
lead flight attendant over the public address system. The hijackers did not allow further messages. According
to the accident report, instructions were only given in English to the passengers from over 30 different nations.
Thus, communication and language difficulties might have affected the emergency preparation. The first officer
and crew assisted the passengers with donning on and handling life vests and showed how to assume the brace
position during impact. Also all loose items were stowed in their appropriate place. Whether the preparations
for ditching had been fully and correctly carried out could not be determined by the accident investigators.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 6/19
The disintegration of the fuselage into four parts resulted in a rapid water ingress, but also enabled a rapid
escape through the openings of the broken fuselage. Almost impossible was the escape for the passengers in
the upside down positioned center section. Here about 50 people were fatally trapped. Apart from this, no
clear pattern of injury related to seating position can be concluded from the accident report, Fig. 4. Among
the surviving passengers the injuries ranged from minor to very serious. The injuries were mainly caused from
impact but some also from drowning. Immediate rescue and first aid was given by people from the beach and
the nearby hotel, soon joined and organized by official forces.
A critical review of the accident report reveals that, apart form the information that the flaps were in retracted
position, no further investigation and discussion into the reason for this is given. The flap position directly
correlates with the approach speed consequently determining the kinetic impact energy. The maximum lift
coefficient based on the wing area for a B-767 is given by Brune and McMasters (1990) with CLmax,flaps = 2.45
for full flaps configuration. The order of the maximum lift coefficient for the clean configuration, i.e. no flaps
deployed, may be reasonably assumed as CLmax,clean 1.2. Taking further the ratio of maximum lift
coefficients as equal to the respective ratio for actual flyable lift coefficients, the ratio of necessary approach
speeds to achieve the same lift may be calculated as:
V2
flaps
V2
clean
=CLclean
CLflaps
1.2
2.45 0.5 (1)
The kinetic energy is also proportional to the velocity squared. Thus, the kinetic energy at impact would have
been in the order of 50% lower for the aircraft in full-flaps configuration compared to the aircraft ditching in
clean configuration. In full-flaps configuration the aircraft could have been flying at about 70% of the speed
required to give the same lift in clean configuration.
3.1.2 Summary from ditching point of view
Based on the information provided by the accident report, this water impact accident has to be rated as planned
ditching. The pilot and to some extend also the crew and passengers were prepared for the water impact. The
accident may also be ranked as an intentionally forced unsurvivable water-impact accident as the hijackers made
improper control inputs and disagreed with the intention to perform an emergency landing.
Key factors regarding the circumstances of the ditching can be split into factors contributing to an at least partly
“successful” ditching and contradictory factors. Key factors positively affecting the outcome of this planned
ditching were:
The low weight due to empty fuel tanks reduced the impact energy.
The landing gear was retracted following general ditching procedures.
Wind and seaway were moderate and consequently had no impact on the ditching situation.
The daylight and the closeness of the crash site to an occupied beach eased evacuation, rescue, and first-aid
given to survivors.
The environmental conditions regarding air and water temperature opposed no additional immediate
threat to the survivors and rescuers.
On the other hand the following key factors are responsible for the high fatality rate (71% of the persons on
board were fataly injured) of this planned ditching:
As the flaps were not set, the airspeed at impact was comparatively high and increased the kinetic energy
at impact.
Pilot and first officer only late (if at all) assumed complete control over the aircraft. Thus, they might
not have been able to carry out a maneuver close to the ditching instructions.
There was no vertical speed indication for the pilots to carry out the ditching maneuver.
The roll angle to the left and the low pitch attitude were opposed to general ditching procedures and
probably increased accelerations and loads at impact.
The impact of the aircraft not only with the water surface, but also with a reef increased loads and
accelerations at impact.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 7/19
Accelerations and disintegration of the aircraft at impact to the water surface and the reef caused imme-
diate fatal injuries to some of the persons on board the aircraft.
The disintegration of the fuselage in four parts lead to immediate flooding of the aircraft. Especially in
the center part of the fuselage being turned up-side-down passengers died from drowning.
The pilots were not able to inform and prepare the cabin crew and the passengers for the ditching situation.
The pilots had no guidance or training for an all-engine-out ditching.
Summing up, the at least partly “successful” outcome of the ditching was basically positively effected by the
conditions at the crash site, while the circumstances on board the aircraft, the ditching maneuver itself and the
preparation of crew and passengers seriously interfered with the survival chance of the persons on board the
aircraft in this planned ditching.
3.2 Unplanned ditching of Boeing 707 in 2000 in Lake Victoria
On 3 Feb. 2000 at 17.36 UTC (20.36 local time) the Boeing 707-351C with registration ST-APY operated by
Trans Arabian Transport (TAAT) ditched unplanned in Lake Victoria about 4km short of runway 12 of Mwanza
Airport, Tanzania. The aircraft was carrying a crew of five all rescued by fishermen.
News of the accident and photos of the damaged aircraft are compiled by the Flight International (2000a,b)
and Sparaco (2000). Much of the accident description presented here is based on the report by Nyamwihura
(2000) and the AIB Bulletin No. 1/00 (2000). The here performed analysis of the flight data recorder (FDR)
parameters is based on the flight recorder replay report by Evans (2000).
Figure 5: Boeing 707 ST-APY afloat more than 12 hours after unplanned ditching in Lake Victoria in 2000.
Missing engines, flaps, and ailerons and damaged left wing tip (left photo). Damaged lower forward fuselage
structure in area of torn away nose landing gear (right photo). Photos: Nyamwihura (2006)
The Boeing 707-351C is based on the design of the first commercially successful four engine passenger jet airliner.
The B-707-351C was designed as a convertible passenger/freighter version with a fuselage length of 46.6m, a
diameter of 3.76m, and a span of 44.4m. The aircraft features a low wing design with conventional tail unit.
The four turbofan engines were mounted below the wing. At the time of the ditching the aircraft operated as
cargo aircraft.
3.2.1 History of flight
The aircraft was operating a charter flight from Khartoum, Sudan to Mwanza, Tanzania for the purpose of
uplifting some cargo of fish fillet bound for Brussels. Mwanza airport, elevation 3763ft has one runway 12/301
which is 3300m long and 45m wide. The crew contacted Mwanza Tower at 16.58 and were advised that there
was no power at Mwanza Airport but efforts were being made to use a standby generator.
After the generator became on line and the runway was lighted, the aircraft commenced a visual approach to
runway 12. At 17.28 the captain reported turning final runway 12 and a landing clearance was given. The
1“runway 12” designates runway direction for take-offs and landings with approximate heading 120and “runway 30” with
heading 300respectively
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 8/19
aircraft made a normal visual approach to runway 12. When it came close to the point were it was expected the
touch down it was observed to overshoot after which it made a left turn. The controller advised the commander
to try runway 30 since the wind was calm and runway 30 had approach lights. However, the commander chose
to use runway 12 again. The accident report only makes explicit reference to the existence of approach lights
for runway 30. As the approach to runway 12 is over the lake and the landing threshold close to the lakeshore,
there is probably no approach light system for this runway.
The aircraft was then observed to proceed on a left base2for runway 12, executed a left base turn3for runway
12 but could not come out aligned properly and therefore proceeded to continue for a wider right base turn for
runway 12. According to the controller, he cautioned the crew of the height, as the aircraft descended very low
over the water level while it was still far from the threshold of runway 12. The aircraft was seen falling into the
lake with a bang moments later. The commander later reported that whilst executing another approach and
with the runway ahead and slightly to the left the aircraft suddenly hit something. There was a lot of noise and
the aircraft came to a stand still. The crew shut down the engines and fired the fire bottles.
Fishermen, who were in the lake, witnessed that the aircraft was making a turn and reported two loud bangs.
They sailed in the direction of the plane, assisted by the SOS flash messages from crew torches. Whereas two of
the crew members were found swimming to the shore, the remaining three were found standing on the aircraft
wing. They all wore life jackets. Apart from fuel leakage into the lake, no other third party damages or injuries
were reported.
On the following day some 18 hours after the accident the aircraft had drifted further in the waters. It was
floating freely with the tail section partly submerged, but the forward fuselage section was standing at an angle,
clear of water. The aircraft was later towed to a position close to the shore where it was de-fueled.
The investigation of the wreckage revealed that both pilots had set the correct QNH = 1015hPa reading on their
pressure altimeters. However, the altimeter on the (left) Pilot seat stopped at 4100ft while that in the (right)
First Officer seat stopped at the correct elevation of lake Victoria, 3720ft. Since the radio altimeter could not
be read at the time of the accident, this disagreement in the altimeter readings could well be a factor in the
accident.
After the accident the captain reported that he took over the controls after the first officer had failed twice to
attain the correct height and turning radius for proper landing. However, the captain handed over controls and
“blacked out” at the time when the aircraft was descending and it was too late to initiate a climb. The captain
himself admitted that this was a “human factors accident”. He attributed his dizziness and the subsequent
blacking out to the fact that he had not eaten food since the evening of the previous day.
3.2.2 Analysis of flight data recorder parameters
The flight data recorder of the aircraft was recovered and analyzed at the Air Accident Investigation Branch,
Aldershot, United Kingdom. The “Flight Recorder Replay Report” by Evans (2000) gives a summary of the
conditions of the Flight Data Recorder (FDR) and the Cockpit Voice Recorder (CVR). Both tapes exhibited
damages typical of that due to water immersion, especially because the water was allowed to dry out gradually.
The parameters recorded by the FDR are: pressure altitude, magnetic heading, normal acceleration, indicated
airspeed, pitch and roll attitude and a FDR sequence number which increments every 10s. All 25 hours of data
from the FDR were recovered. The data quality is described as satisfactory with a few data losses throughout
the flight. However, due to the tape damage in the range of the final few seconds of the recording, there were
many more relevant data errors.
The data of the final 20min from the FDR given in tabular form in the appendix of the report were digitized
for further analysis here. From the magnetic heading and the indicated airspeed an approximate track of the
aircraft is derived by simple time integration of the data. Obviously faulty data were replaced by linearly
interpolated values. No corrections were made to the indicated airspeed, used to derive the ground speed,
and to the magnetic heading, used to derive the true heading. Thus the airplot given in Fig. 6 gives only
an approximate impression of the aircraft motion within the final 20min. The position of the aircraft in the
horizontal plane is given with respect to the final position of the aircraft. The runway 12/30 of 3300m at
Mwanza Airport is plotted at an estimated position where the aircraft could have landed in the first approach.
This position of the runway corresponds to the reported impact location of about 4km from the end of the
runway.
2“left base”: part of approach pattern flying square to the runway located to the left
3“left base turn”: part of approach pattern turning left for the final approach
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 9/19
−15 −10 −5 0 5 10 15 20
2
4
6
8
10
12
14
16
18
position with respect to final position in direction of runway, 103m
altitude, 103ft
Figure 6: Final 20min track of Boeing 707 ST-APY unplanned ditching in Lake Victoria derived from FDR
readout: air plot with respect to final aircraft position (left) and altitude profile calibrated with QNH = 1015hPa
with respect to runway orientation (right). Black dots indicates final aircraft position and solid straight lines
approximately mark Mwanza airport runway 12/30.
Corresponding to the airplot an altitude profile of the aircraft position was calculated. The pressure altitude from
the FDR was calibrated according to the International Standard Atmosphere and formula by the International
Civil Aviation Organization (ICAO) (1993) with the help of the reported QNH = 1015hPa. This calibration
does not account for the deviation from the standard temperature model. Together with the uncertainties of
the pressure measurements this is the probable cause for the altitude after calibration being partly lower than
the elevation of 3763ft of the runway of Mwanza airport. The altitude profile corresponding to the air plot is
plotted in Fig. 6 with respect to the orientation of the runway 12/30. Despite the uncertainties the altitude
profile clearly shows the first approach and the corresponding overshoot near the threshold of the runway. The
plot also reveals the constantly decreasing altitude after the aborted second approach leading to the unplanned
ditching. All in all the airplot together with the height profile visualizes the above reported flight path of the
aircraft.
In order to determine the flight parameters at the time of impact, the final 30s of the FDR data are evaluated
and plotted (see Fig. 7). The time scale is given with respect to the end of the recording. The symbols mark
those time steps where data is assumed to be reliable. The rate of change of the magnetic heading together
with the final roll angle of about 13to the left reveals, that in the final seconds the aircraft was changing from
right turn to left turn for the final approach to Mwanza Airport. The final altitude of 3767ft is very close to
the elevation of Mwanza airport of 3763ft. Regression analysis of the final seconds of the altitude over 10s to
20s gives sinking rates of 5ft/s to 9ft/s. The final indicated airspeed was 123kts and the pitch attitude about
3.8. The parameters of the final seconds are reasonably assumed to be the impact conditions of the aircraft on
the water.
3.2.3 Damages to aircraft structure
The reported damage to the aircraft which is noticeable from photos (see Fig. 5) of the wreckage, reads as
follows:
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 10/19
−30 −25 −20 −15 −10 −5 0
3600
3800
4000
altitude, ft
−30 −25 −20 −15 −10 −5 0
100
150
200
heading, deg
−30 −25 −20 −15 −10 −5 0
120
130
140
airspeed, kts
−30 −25 −20 −15 −10 −5 0
−50
0
50
roll angle, deg
−30 −25 −20 −15 −10 −5 0
−10
0
10
time before end of FDR recording, s
pitch angle, deg
Figure 7: Final 30s of FDR parameters readout of Boeing 707 ST-APY unplanned ditching in Lake Victoria.
All four engines and their pylons separated on impact each taking with it about 2.4m of the wing structure
forward of the front spar. There was no indication of any engine colliding with any part of the remaining
aircraft structure after separation.
The left wing showed severe impact damage towards the tip (see left photo of Fig. 5). All spars and skins
showed 45fractures indicating that they had failed under heavy impact loading. The damage to the
right wing was relatively less severe than of the left wing. The ailerons of both wings had been lost. But
most of the flaps which were in landing configuration were in place, though damaged to varying degrees.
The nose landing gear was ripped of at impact.
The fuselage showed extensive damage to the underside structure. The damage starts from the nose
landing gear and extends 3m backwards and to the left of the center-line (see right photo of Fig. 5). The
observation goes along with the FDR evaluation of the aircraft being in a left turn with roll angle to the
left and is also consistent with the damage to the left wing. All the equipment in the electronics bay fell
into the lake when the lower fuselage was torn open by the separating nose landing gear.
There was no fire.
Apart from the separation of parts of the aircraft and local damages to the under fuselage no global
damages to aircraft with respect to the global structure is apparent. No failure or buckling of the upper
shell of the fuselage was reported or is visible in the photos of the damaged aircraft.
3.2.4 Analysis of accident from ditching point of view
The key factors for surviving a ditching event are the approach conditions, the configuration of the aircraft, the
environmental conditions and the circumstances for the rescue.
There were several positive key factors contributing to this “successful” unplanned ditching in Lake Victoria:
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 11/19
The aircraft weight was relatively low, i.e. well below the certified maximum landing weight of 122t
(Department of Transportation Federal Aviation Administration, 1984), as the aircraft was supposed to
pick up cargo at Mwanza airport and had burned some fuel during the flight.
With flaps set to landing configuration, the final airspeed of about 123kts was close to a normal landing
speed.
The rate of descent of the aircraft of about 5ft/s to 9ft/s was in the order for a normal landing or ditching.
The aircraft had a nose-up pitch attitude of about 3.8. That is favorable, although it is by far smaller
than the pitch attitude of 12recommended in the ditching investigations of this aircraft type (Thompson,
1955).
The moderate damages to aircraft caused no direct threat to the health of the crew and enabled a long
floatation time.
There were no indications of wind, thus it is reasonable to assume that the aircraft ditched into calm
water.
Immediate rescue of the crew was possible due to the near-shore location of the impact.
The reported air temperature of 25C as well as the water temperature in the order of 25C (Mangeni and
Ngirane-Katashaya, 2005) constituted no immediate threat to the health of the crew waiting for rescue or
swimming to the shore.
But also some key factors threatened the “successful” outcome of this unplanned ditching:
The large 13roll angle of the aircraft lead to an asymmetrical impact with the water surface. Although
the left wing apparently hit the water first, the aircraft must have leveled during the landing phase, as
damages occurred also on the right side of the aircraft to wing and fuselage. As reported, also the right
engines with their pylons were ripped off, thus they must have come to water at a phase of the landing
with a significant forward speed.
The landing gear was extended for the landing at Mwanza airport, which is opposed the recommended
procedure for a planned ditching for this aircraft type (Thompson, 1959).
The accident occurred during darkness, usually making rescue operations more difficult.
As the aircraft apparently did not sink, all crew members should have awaited rescue on the aircraft wing
instead of taking the risk of trying to swim to the shore.
Thus it may be concluded from this accident that an aircraft designed with respect to planned ditching also has
a good chance to survive an unplanned ditching under circumstances close to the planned ditching conditions.
Apparently even ditchings with a significant roll angle are not unsurvivable as the aircraft levels itself during
the landing phase. Probable reason is the partly planing on the engines which reduces the sensitivity to a role
angle.
3.2.5 Comparison of ditching tests for this aircraft type to the unplanned ditching accident
The model ditching experiments conducted for the Boeing 707 by Thompson (1955) considered pitch attitudes
of 12, 9and 6resulting in landing speeds of 100kts, 104kts and 114kts respectively for an aircraft weight
of 59t in full scale and flaps set to 50landing configuration. Also tests were carried out with no flaps for
the respectively full scale landing speeds of 119kts, 127kts and 146kts. Flaps and engines were mounted with
scaled strength, and the fuselage bottom was investigated in undamaged and scaled strength, conditions. The
conclusion from these tests was, that “the aircraft should be ditched in a nose-high attitude of about 12with
landing flaps down”, because of the lower landing speed going along with this condition leading to less fuselage
damages. It was further anticipated that “the engine nacelles will probably be torn away and the fuselage
bottom will most likely be damaged enough to cause rather rapid flooding”. Also the failure of flaps was noted
in the tests.
Further ditching tests by Thompson (1959) in calm water and head waves were undertaken for the Boeing 707
with the favored pitch attitude of 12at a full scale landing speed of about 120kts necessary for a full scale
aircraft weight of 116t and flaps set to 50landing configuration. The ditching test with landing gear extended
and retracted were carried out. For extended nose landing gear it was noticed, that it always was torn off in
calm- or rough-water ditchings. The test also showed, that “there was no appreciable difference in behavior or
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 12/19
damage to the fuselage bottom whether the model was ditched with the nose gear extended or retracted”. For
the main landing gear the conclusion was: “it appears that, if the main gear fails, there is little choice between
ditching with the gear extended or retracted; but, if the main gear does not fail, a more dangerous behavior
results. Therefore, it is recommended that the gear remain retracted in a ditching”.
The landing speed as dominant parameter in ditching in the above ditching tests were almost the same as in
this ditching accident, whereas the pitch attitude was considerably lower in the ditching accident. The failure of
all four engines, some of the flaps and ailerons and the damage to the bottom of the fuselage in this unplanned
ditching go along with the above cited conclusions of the ditching tests. Also the failure of the nose landing gear
was predicted by the tests. However, the good floatation capability of the aircraft, even in damaged condition,
was not foreseen.
3.3 Planned ditching of Boeing 737 in 2002 in Bengawan Solo River
On 16 Jan. 2002 at about 09.24 UTC (16.24 local time) the Boeing 737-300 with registration PK-GWA operated
by PT Garuda Indonesia as flight GA421 ditched in the Bengawan Solo River, Central Java, Indonesia. The
aircraft had suffered a dual engine flame out attributed to excessive hail and rain ingestion. At least two relight
attempts on both engines failed and the aircraft suffered run-out of electrical power. There were 6 crew members
and 54 passengers on board the aircraft. One cabin crew member was fatally injured in addition to one other
cabin crew member and 12 passengers which were seriously injured.
The Boeing 737-300 is a short-to-medium-range, single-aisle, narrow-body aircraft with a length of 33.40m and
a span of 28.88m. The fuselage has a height of 4.01m and a width of 3.76m. The aircraft has a low wing and
a conventional tail design. Propulsion is provided by two under-wing turbofan engines. The ditched aircraft
was configured with 104 passenger seats: 22 in business class and 82 in economy class. An aft-facing, double-
occupancy flight attendant jump-seat was located by the forward passenger door and single occupancy flight
attendant jump seats were mounted behind the left-hand and right-hand aft lavatory.
The accident investigation was carried out and reported by the National Transportation Safety Committee -
Department of Communications - Republic of Indonesia (2006). Assistance was provided by the National Trans-
portation Safety Board (NTSB) (2005) leading to own safety recommendations. Aircraft, aircraft component
and engine manufacturer participated in the investigation. The DFDR and CVR readout was performed by the
Air Accident Investigation Branch (AAIB), Farnborough, UK. The accident was only briefly reported in the
international press, e.g. Ionides and Learmount (2002).
3.3.1 History of flight
En-route from Ampenan to Yogyakarta, Indonesia the B-737 flight entered an area covered by Cumulonimbus
cells while descending from cruise altitude at FL310 to FL190. The crew prepared the aircraft and passengers to
enter turbulence. Both engines flamed out in severe turbulence and heavy precipitation. Total electrical power
loss was noted by the flight crew after at least two engine restart attempts and one attempt of APU start.
The engine flame-out has been attributed “to excessive water/hail ingestion on the engines which is beyond the
engine certified capabilities.” The relight attempts were unsuccessful since the aircraft was still in precipitation
beyond the engines’ certified capabilities. The power loss was due to the battery inability to maintain sufficient
power because of inadequate battery maintenance procedures. Because of the total power loss CVR and DFDR
immediately ceased recording.
The pilot spotted the Bengawan Solo River from about 8000ft altitude and decided the ditch the aircraft on
the river. Hence, this accident is a planned ditching. The flight crew is reported to have “announced to the
flight attendant to prepare emergency landing procedure”. The passengers were already advised to wear seat
belts before because of the flight through severe turbulence. One flight attendant is reported to “not expect
to ditch on a river”. It is not given in the accident report whether and which parts of the ditching procedures
were followed by the flight and cabin crew.
The ditching was executed into the Bengawan Solo River in upstream direction between two iron bridges located
approximately 2km apart. At the ditching site the smooth flowing river is approximately 75m wide and 1 to
5m deep. The aircraft settled down on its belly in partly submerged position (Fig. 8).
Key parameters of the ditching relevant water-impact condition are the aircraft’s weight, speed, and orientation
with respect to the water surface. Due to the unavailability of DFDR data for the water impact, the impact
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 13/19
Figure 8: Boeing B-737 PK-GWA resting in Bengawan Solo River, Indonesia after ditching in 2002 with no
visible damages to the primary overall aircraft structure apart from the distorted engine mountings. Photos
(left, right): Indra (2002), Lim (2005).
conditions can only be estimated from other data and information. Based on the damage pattern the aircraft
presumably impacted the river in nose-up attitude.
The flaps and slats were in retracted position when ditching and were found largely intact. This clean configu-
ration at ditching gives rise to a higher approach speed compared to normal landings with flaps. The maximum
lift coefficient at landing for the B-737-300 is given by Brady (2007) to CLmax, landing = 2.88 and from the
data given by Englar (1996) the maximum lift coefficient in clean configuration of the B-737 is estimated to
about CLmax, clean = 1.25. Following the approach presented in section 3.1 expressing the relation of lift
coefficient to speed (1), the kinetic energy at landing with flaps would have been about 43% of that at landing
in clean configuration. The approach speed could have been in full flaps configuration only 66% of that in clean
configuration.
While the intact right-wing fuel tank was found with 2800 liters of fuel, the leaking left-wing fuel tank was found
with 400 liters. No information on the filling of the center fuel tank is given in the accident report. Assuming
that both wing tanks carried the same amount of fuel and the center tank was almost empty at impact, the
aircraft had about 4.5t of fuel on board which is about 1/4 of the fuel capacity. With only slightly more than
half of the passenger seats being occupied the weight of the aircraft at impact can be estimated to be in the
order of 80% of the aircrafts maximum landing weight.
3.3.2 Analysis of damages
The primary overall aircraft structure appeared undamaged apart from distorted engine mountings (Fig. 8).
The left engine is reported being still attached to the pylon, while the right engine was detached from its pylon.
As both engines were damaged to the same degree, the aircraft likely impacted the water without a considerable
roll angle.
The nose landing gear and the right hand main landing gear were found detached from their attachments
approximately 50m from the aircraft final rest. The wings and control surfaces were largely intact and the
leading edge of the horizontal stabilizer suffered light water impact damages. The auxiliary power unit (APU)
and battery as well as the CVR detached from the aircraft and were found 200-300m downstream of the aircraft.
The fuselage was heavily damaged at the bottom aft of the right hand cargo door. Here a hole opened which
was large enough for the flight attendants seated in the rear of the aircraft to be pulled out of the plane at
water impact. The upper fuselage shell in the aft just in front of the vertical stabilizer showed skin buckling.
The forward part of the fuselage was also heavily damaged between the cockpit section and the front wing spar
bulkhead (see Fig. 8 left photo). The accident report clearly attributes the damages of the aft fuselage section
to the impact in the ditching with a pitch-up attitude. The buckling of the aft upper shell can be assumed to
be related to the upwards bending moment introduced at impact of the aft fuselage section.
The damages reported and depicted in photos of the interior of the aircraft (Fig. 9) match those described above
for the overall aircraft structure. The largest damages occurred in the aft section of the fuselage where the left
and right hand lavatory and the lower part of the galley were ripped apart from the aircraft. The floor in this
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 14/19
section of the aircraft collapsed and opened a large hole. In the front part, the floor aft of the cockpit door
buckled. At the second row of the business class, the floor on the aisle collapsed. At least partly related to the
disintegration of the floor is the detachment of seats. In line with the fuselage deformation, some ceiling and
wall panels separated.
(1)
(13)
(46)
Figure 9: Boeing B-737 PK-GWA ditching in 2002 in Bengawan Solo River, Indonesia: Overview of interior
damages and occupant injuries (left). Damages to floor, seats, and wall panels of business class (upper photo)
and to ceiling panels, aft lavatories, and aft galley (lower photo). Graphic: Accident report by National
Transportation Safety Committee - Department of Communications - Republic of Indonesia (2006). Photos:
Moch (2008).
The B-737-300 has 6 exits: on the left hand side the forward and aft passenger doors, on the right hand side
the forward galley door and the aft service door, and the left and right hand overwing exits. Both aft doors
were unusable due to structural deformation of the fuselage and also because of the adjacent floor collapse. The
forward passenger door was jammed closed as well. Thus only half of the installed exits were usable for the
evacuation. The cockpit door was only slightly jammed and could be kicked open by the flight crew.
3.3.3 Evacuation and injuries
The two uninjured flight attendants seated in the front part were able to open the forward galley door on right
hand side. This was the only immediately available exit. The respective escape slide inflated automatically. The
uninjured flight crew assisted in organizing the passenger evacuation. The other two flight attendants seated
in the aft were pulled out of the plane during the ditching. One sustained serious injuries and the other fatal
injuries attributed to the impact of the aircraft with the water and the ground of the river.
As depicted in Fig. 9 of the 6 passengers seated in the business class 5 were seriously injured. These injuries
can be correlated to the damages in the front part of the fuselage, especially the disintegration of seats. There
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 15/19
is no obvious pattern regarding the distribution of the 7 seriously injured passengers among the 48 passengers
seated in the economy class.
Due to the total power loss the flight crew was not able to inform the air traffic control of the emergency
situation and the decision to ditch the aircraft. With the help of his mobile phone the pilot notified the Garuda
Operations Centre in Jakarta.
Right after the ditching, a fisherman on the river offered his help. With his assistance from the outside the
right-hand side overwing exit could be opened. The seriously injured business-class passengers were evacuated
through the forward galley door assisted by the flight attendants. The economy-class passengers were evacuated
through the right-hand side overwing exit. Some passengers did not leave the aircraft immediately; instead they
tried to retrieve their personal belongings stowed in the overhead bins. Villagers from around the accident
site helped with the evacuation, temporarily sheltering, and the transportation of seriously injured persons to
nearby hospitals. The evacuation process of all passengers to save locations was completed approximately one
hour after the ditching. Official and professional personal arrived at the crash site hours after the crash.
3.3.4 Analysis of accident from ditching point of view
This planned ditching with only one fatality out of 60 persons on board shows that modern type aircraft can
be successfully ditched. Contributing key factors to the positive outcome were:
The aircraft impacted the water in a nose-up attitude.
The aircraft had a moderate weight during the water impact.
The level of the loads and accelerations at water impact were within a survivable limit regrading injuries
to the persons on board and damages to the aircraft.
Due to the water depth at the ditching site the floatation capability of the aircraft was not an issue. The
usage of water survival equipment was not necessary as well.
Third persons provided almost immediate help assisting in the evacuation, conducting first-aid treatment,
lodging the rescued passengers, and transporting the injured persons to nearby hospitals.
The weather and daylight eased the execution of the ditching and the evacuation of the persons on board
the aircraft.
On the other hand there were also factors hampering a successful outcome of this ditching:
There were only a couple of minutes between total engine flame-out and ditching leaving only a short time
for the flight crew to inform the cabin crew and passengers and to prepare the aircraft for ditching.
The approach speed at ditching without slats and flaps was considerably higher than the respective landing
speed with slats and flaps.
The damages to the fuselage shell would have caused rapid flooding leading to significant reduction of
floatation time in case the aircraft came to rest in deeper water.
The detachment of seats very likely contributed to the injuries of the passengers.
The immediate availability of only one exit could have been a possibly fatal trap for the persons on board
in case of necessarily quick evacuation because of rapid flooding or fire and smoke.
The breakage and collapse of the floor likely caused additional injuries and slowed the evacuation process.
The arrival of professional search and rescue forces took hours, thus leaving the evacuation and first-aid
injury treatment to local private persons.
This planned ditching was obviously well executed ensuring the overall integrity of the aircraft structure being
responsible for the almost totally successful ditching regarding the fatality rate. On the other hand, the failure
of subsystems such as seats and parts of the floor and the unavailability of exits could have led under slightly
different boundary conditions to a considerably more fatal outcome of this ditching event.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 16/19
4 Assessment of planned and unplanned ditching events
The listed planned ditching accidents (Table 1) show that this type of accident is rare but does exist. The
survival rate for the individual accidents range from completely successful to totally fatal. The damages range
from minor local damages to the shell structure of the fuselage to complete breakup of the fuselage in several
pieces and massive local damages and rapture. An average of one ditching in six years can be deduced for
passenger transport jet aircraft planned ditchings. Further statistical evaluation of the number of water impact
accidents would have to be conducted taking into account the constantly increasing over-water flight operations.
Even without these detailed statistics it is evident, that there is no increase in ditching events as predicted by
earlier water-impact accident studies. This goes along with the constantly increasing safety of air transport.
In order to focus on passenger transport jet aircraft, the present study does not consider the planned ditching on
6 Aug. 2005 of the twin-turboprop ATR-72-202 with registration TS-LBB operated by Tuninter. The high-wing
aircraft suffered a dual engine flame-out after running out of fuel and ditched in the Mediterranean. Of the 4
crew members and 35 passengers on board, one cabin crew member and 15 passengers suffered fatal injuries and
the remaining occupants sustained minor to serious injuries. The fuselage broke fore and aft of the wing section
into three parts. This event shows remarkable similarities to the accidents analyzed here. Together with the
findings from the accident investigation into the Airbus A320 ditching on 15 Jan. 2009 into the Hudson River
these accidents can supplement the presented case studies.
The compiled lists of actual ditchings provide some aspects which should be addressed within the design and
certification of an aircraft. All planned ditchings of passenger transport jet aircraft listed above (Table 1) as
well as those of the ATR-72 in 2005 and the A320 in 2009 had to be carried out with all engines out. Despite all
safety measures concerning the reliable operation of the engines, also the recent ditchings show that there are
circumstances leading to an all-engines-out situation and the necessity to ditch the aircraft. Along with the loss
of engine thrust other systems important for the ditching maneuver such as flaps and navigational instruments
may fail. These systems are directly linked to the speed and thus energy at water impact, and to the possibility
to reduce this energy during the final approach. Hence, this possibility should be addressed within the ditching
investigation during design and certification of aircraft. Also the ditching procedures and training given to the
flight crew should cover this issue (see reports on ditchings of B-767 in 1996 and ATR-72 in 2005), because the
skill and performance of the flight crew is a key factor in a ditching maneuver.
A crucial aspect already marked by earlier water-impact accident studies is the communication on board of the
aircraft in the preparation phase for the ditching. As the ditchings of the B-767 in 1996 and of the ATR-72 in
2005 show, it is important that the flight crew informs the cabin crew and the passengers. Additionally, both
ditchings point at the necessity of the cabin crew to inform the passengers about the situation and assist with
seat-belts and life-vests.
As the three case studies show, the structural integrity and the fatality in ditching accidents are directly
correlated. Common reason is the state of motion including the kinetic energy of the aircraft at water impact.
The accelerations during water impact are responsible for impact injuries to the persons on board. For the
global loads acting on the aircraft the overall motion of the aircraft is decisive. Further fatalities are caused
if the fuselage is disrupted due to excessive loads. The rapid flooding in case of disintegration of the fuselage
structure is reason for further fatalities. Thus, the most important issue in ditching is to minimize the kinetic
energy. The motion of the aircraft and the related global forces are to be investigated within studies of aircraft
ditching by tests and numerical simulations. Despite large differences concerning circumstances and type of
aircraft, the ditchings of the B-767 in 1996 and the ATR-72 in 2005 show remarkable similarities regarding
accelerations and global loads exceeding the fuselage strength. This can be concluded from the type of injuries
and the fuselage breakage pattern.
Loads on smaller parts of the aircraft such as engines, flaps, and landing gear, including their possible separation,
have to be accounted for as well if these loads effect the motion of the aircraft substantially. As the unplanned
ditching of the B-707 in 2000 shows, local damages to the shell of the aircraft are of secondary influence to the
success of a ditching. Despite local damages the floatation time was not critically reduced in this case.
The presented analysis of the ditching accidents reveals that not only the overall integrity of the aircraft structure
is decisive regarding survivability and injuries, but also the crashworthiness of interior subsystems such as for
example seats, seat-belts, floors, hatracks, galleys, and exits. The importance of cabin subsystem design and
integration has been stressed already in earlier ditching studies (National Transportation Safety Board (NTSB),
1972) and remains an issue as the ditchings of the B-737 in 2002 and of the ATR-72 in 2005 show.
Three of the six planned ditchings were carried out in rivers with the advantage of quick rescue possibilities.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 17/19
Additionally, the importance of aircraft floatation and personal floatation devices is reduced.
Two out of the six planned ditchings (B-767 in 1996 and the B-737 in 2002) have the impact with obstacles
combined with the water impact in common with overrun water impact accidents. The impact with obstacles
in the water increase the deceleration of the aircraft and increase the damages to the aircraft partly up to
the disintegration of the fuselage structure. Thus, this type of accident directly poses additional threats to
the persons on board. Additionally, the damages caused by impact with obstacles considerably increase water
ingress and consequently reduce floatation time.
The presented case study of the B-707 unplanned ditching in 2000 is representative for a typical controlled,
although unplanned, water contact during final approach. Examples of similar situations and similar outcome
exist, e.g. unplanned ditching of the B-727 in 1978 investigated by the National Transportation Safety Board
(NTSB) (1978). The reason for these unplanned ditchings being generally so successful is the fact that during
normal final approach the speed is reduced corresponding to the flap setting, the sink rate is moderate, and
the weight is reduced by the amount of fuel burned during the flight. Hence, if the aircraft has a nose-up pitch
attitude these conditions are close to the generally recommended approach conditions for ditching. This fact
supports the justification that the investigation and design of aircraft with respect to planned ditching also
cover likely unplanned ditching situations.
An extraordinary fact about the B-707 ditching in 2000 is that the unplanned ditching was successful despite
a considerable roll angle. The conclusion is that the roll angle was, in this case, of minor importance to the
overall ditching and does not necessarily lead to fatal ditching events.
5 Conclusion
As one pillar within the framework of the “triangle of aircraft ditching investigation methods” this paper focuses
on the “Ditching Accident Analysis”. The analysis consists of the compilation of key facts of planned ditching
events of passenger transport jet aircraft and the presentation of three ditching accident case studies.
The overall absolute and relative small number of planned ditching accidents of passenger transport jet aircraft
may raise the question, how much the aircraft design and operation procedures should account for this situation.
As, within the certification of an aircraft, ditching is one aspect (paragraph “25.801 Ditching” by the European
Aviation Safety Agency (EASA) and the Federal Aviation Administration (FAA)), the overall question whether
it is worthwhile to account for ditching at all can not be raised unless challenging this regulation. Although
the ditching regulations basically focus on the planned ditching scenario, the intention is also to cover, at least
partly, the unplanned ditching case. As the number of unplanned ditchings is higher, this increases the value
of covering ditching within the design, certification and operation of aircraft. All in all the aspect of ditching
should not be over-emphasized compared to more frequently occurring hazards. Thus, issues solely relevant for
ditching should be balanced within the overall safety introduced in aircraft design and operation.
The distinct feature of aircraft ditching is the transition of loads acting on the airframe and parts of it within the
four ditching phases Fig. 1: inertia loads, aerodynamic loads, hydrodynamic loads and hydrostatic loads. Due
to the density of water the hydrodynamic forces dominate the external forces. Thus, the correct modeling of
the hydrodynamic forces taking into account the high forward velocity is decisive for aircraft ditching modeling.
As the aircraft deforms under the external loads, modeling of aircraft ditching is a fluid-structure interaction
problem. Thus, an adequate and matching level of detail for the modeling of the hydrodynamic loads as well as
the structure dynamics is required. This means that the level of physical modeling as well as the disretization in
space should be selected appropriately. With respect to the aircraft motion and global loads the hydrodynamic
modeling is more important than the structural modeling.
The modeling of local deformations is difficult and the significance with respect to loads and motion of the
aircraft during ditching is not obvious. But force modifying the separation of aircraft components such as
engines and flaps should be modeled. A challenging task is the modeling of the combined water and (soft)
ground impact as seen in two of the ditching studies.
As the ditching event case studies show, there are no clearly defined initial conditions for ditching of aircraft.
Moreover, ditching is influenced by various parameters. Consequently, a trade-off between parameter studies to
cover these parameter envelopes and detailed physical modeling has to be chosen. Detailed physical modeling
requires respective computational effort and thus can be carried out only for a restricted number of cases.
Hence, a combination of applying detailed physical modeling in selected cases for validation purposes and using
reduced order modeling for a variety of practically relevant cases appears desirable.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 18/19
Bibliography
Agenzia Nazionale per la Sicurezza del Volo (ANSV). FINAL REPORT - Accident involved aircraft ATR 72, registration marks
TS-LBB, Ditching off the coast of Capo Gallo (Palermo - Sicily), August 6th 2005, Jan. 2009.
AIB Bulletin No. 1/00. AIB Bulletin - Accident to Boeing 707 - 351C ST-APY in Lake Victoria on 3. Feb. 2000. The United
Rebublic of Tanzania - Ministry of Communications and Transport - Accident Investigation Branch, Dar es Salaam, Tanzania,
2000.
C. Brady. The Boeing 737 technical guide. London : Tech Pilot Services, 2007.
G. W. Brune and J. H. McMasters. Applied Computational Aerodynamics, volume 125 of Progress in Astronautics and Aeronautics,
chapter 10 “Computational Aerodynamics Applied to High Lift Systems”, pages 389–433. American Institute of Aeronautics
and Astronautics (AIAA), Washington D.C., 1990.
CNN interactive. Ethiopian airline crash kills at least 50. http://edition.cnn.com/WORLD/9611/23/hijack.ethiopia.pm/index.
html Last checked: 01.05.2009, Nov. 1996a.
CNN interactive. Bizarre ordeal recounted in ethiopian airlines crash. http://web.archive.org/web/20070205144457/http://www.
cnn.com/WORLD/9611/24/ethiopia.hijack/index.html Last checked: 01.05.2009, Nov. 1996b.
CNN interactive. Honeymooners capture dramatic images of ethiopian jet crash. http://web.archive.org/web/19990117075208/
http:/www.cnn.com/WORLD/9611/26/comoros.crash/index1.html Last checked: 01.05.2009, Nov. 1996c.
D. K. Cosper and G. A. McLean. Analysis of ditching and water survival training programs of major airframe manufacturers and
airlines. Technical Report DOT/FAA/AM-98/19, LB & M Associates and Federal Aviation Administration, July 1998.
Department of Transportation Federal Aviation Administration. Type Certificate Data Sheet No. 4A26, July 1984.
R. J. Englar. Application of pneumatic lift and control surface technology to advanced transport aircraft. In Transportation Beyond
2000: Technologies Needed for Engineering Design, pages 371–397, Feb. 1996.
Ethiopian Civil Aviation Authority - Flight Safety Department. Aircraft Accident Investigation Report Ethiopian Airlines B767
(ET-AIZ) Aircraft Accident In The Federal Islamic Republic Of The Comoros, In The Indian Ocean On November 23, 1996.
Addis Ababa, Ethiopia, May 1998.
European Aviation Safety Agency (EASA). Certification Specifications For Large Aeroplanes CS-25, Amendment 5. Brussels,
Sept. 2008.
A. Evans. Accident to Boeing 707-351C ST-APY near Mwanza Airport on 3rd February 2000. Flight record replay report, Air
Accident Investigation Branch (AAIB), Aug. 2000.
Federal Aviation Administration (FAA). Criminal Acts Against Civil Aviation, 1996.
Federal Aviation Administration (FAA). Federal Aviation Regulation Part 25 Airworthiness Standards: Transport Category
Airplanes, Subpart D - Design And Construction, Emergency Provisions, Sec. 25.801 Ditching, Amendment 25-72, Aug. 1990.
D. E. Fink (Editor). Aircraft search for missing peruvian 727. Aviation Week & Space Technology, 133(12):42, Sept. 1990.
Flight International. Altimeter discrepancies found in crashed 707. Flight International, 157(4717):7, Feb. 2000a.
Flight International. Trans arabian’s 707 splashes down. Flight International, 157(4715):11, Feb. 2000b.
Flight International. Prepare to ditch. Flight International, 86(2892):241, Aug. 1964.
A. Indra. http://www.airdisaster.com/photos/ga421/photo.shtml Last checked 01.05.2009, 2002.
International Civil Aviation Organization (ICAO). Manual of the ICAO standard atmosphere : extended to 80 kilometres (262
500 feet). Montreal, 3rd edition, 1993.
N. Ionides and D. Learmount. Investigators probe sudden engine failure on descending Garuda 737. Flight International, 161
(4815):7, Jan. 2002.
Joint Rescue Coordination Centre (RCC) Halifax. SAR Operations Report for RCC Case 1914, 11 September 1990 (Boeing 727
PLI 976), Sept. 1990.
K. Lim. Both engines of a Boeing 737-300 flamed out and ditched successfully! http://www.geocities.com/khlim777 my/
asditching.htm Last checked: 18.01.2008, 2005.
B. Mangeni and G. Ngirane-Katashaya. Preliminary investigation of Lake Victoria groundwater situation from AVHRR data. In
31st WEDC International Conference, pages 291–294, Kampala, Uganda, 2005.
H. Moch. Pictures of B-737 after ditching in Bengawan Solo River on 16.01.2002. Personal communication, July 2008.
National Transportation Safety Board (NTSB). Safety Recommendation. Washington, D.C., Aug. 2005.
National Transportation Safety Board (NTSB). Aircraft accident report, Overseas National Airways, Inc., Douglas DC-9, NC935F,
Operating as Antilliaanse Luchtvaart, Maatschappij Flight 980, near St. Croix, Virgin Islands, May 2, 1970. Washington, D.C.,
Mar. 1971. NTSB/AAR-71/8.
National Transportation Safety Board (NTSB). Aircraft accident report, National Airlines, Inc., Boeing 727-235, N4744NA,
Escambia Bay, Pensacola, Florida, May 8, 1978. Washington, D.C., Nov. 1978. NTSB/AAR-78/13.
National Transportation Safety Board (NTSB). Special Study, Passenger survival in turbojet ditchings (A critical case review).
Washington, D.C., Apr. 1972. NTSB/AAS-72/2.
National Transportation Safety Board (NTSB). Preliminary Report Aviation. Washington, D.C., Jan. 2009.
National Transportation Safety Committee - Department of Communications - Republic of Indonesia. PT. Garuda In-
donesia GA421; B737-300 PK-GWA; Bengawan Solo River, Serenan Village, Central Java; 16 January 2002, 2006.
KNKT/02.02/06.01.33.
J. Nyamwihura. Accident to Boeing 707 - 351C ST-APY at Mwanza on 3. February 2000. Personal communication, Feb. 2006.
J. Nyamwihura. Preliminary Report - Accident to Boeing 707 - 351C ST-APY at Mwanza on 3. February 2000. The United
Rebublic of Tanzania - Ministry of Communications and Transport - Accident Investigation Branch, Mar. 2000.
A. Patel and R. J. Greenwood. Transport water impact and ditching performance. Technical Report DOT/FAA AR-95/54, FAA
William J. Hughes Technical Center, Atlantic City, N.J., Mar. 1996.
P. Sparaco. Investigators probe african accidents. Aviation Week & Space Technology, 152(10):43, Mar. 2000.
J. Tahliani and M. Muller. Transport water impact part II. Technical Report DOT/FAA AR-95/112, FAA William J. Hughes
Technical Center, Atlantic City, N.J., May 1996.
W. C. Thompson. Model ditching investigation of the Boeing 707 jet transport. Research Memorandum for the Civil Aeronautics
Administration SL55K08, National Advisory Committee for Aeronautics (NACA), Washington, Oct. 1955.
W. C. Thompson. Rough-water ditching investigation of a model of a jet transport with the landing gear extended and with various
ditching aids. Technical Note D-101, National Aeronautics and Space Administration (NASA), Washington, Oct. 1959.
B. W. Townshend. Ditch or crash-land? Privately printed by Norman Starbuck & Co Ltd., Cranleigh, Surrey, 1965.
6th International KRASH Users’ Seminar (IKUS6), 15-17 June 2009, Stuttgart 19/19
... This event is described as planned and controlled aircraft landing on free-surface due to extreme emergencies, e.g., bird strike, thunderstorm weather, fuel exhaustion, crew error, terrorist hijack, mechanical failure, etc. The crew, with the aircraft under control, deliberately and gently lands on water according to recommended procedures given in the flight manual [1]. It is categorized into four stages [2]: approaching, impacting, landing and floating, among which, the impacting and landing are the fatal stages during this process and result in complex fluid-structure interactions [3]. ...
... The previous document [4] shows that there are many tragic aircraft crash accidents within 219 shocked emergency ditching since 1938. Several well-known airline plane ditching attempts have been analyzed [1,5]. Additionally, helicopter crash mishaps occurring due to multiple reasons. ...
Article
Full-text available
Aircraft, such as amphibious planes, airliners, helicopters and re-entry capsules, are frequently subject to impacting loads from water-landing/ditching on various free surfaces, especially under wave conditions. Understanding and quantifying the water-landing/ditching performance on wave surfaces are of fundamental important for the design and certification of crashworthiness in the field of aerospace engineering. This study aims to numerically assess the effect of wave surface on water-landing process of an amphibious aircraft. The numerical implementation is realized in Reynolds-averaged Navier–Stokes (RANS) framework by combining finite volume method (FVM), volume of fluid (VOF) approach and velocity-inlet wavemaker. The temporal-spatial characteristic of numerical wave and the accuracy of presented model are, respectively, validated by analytical wave and convergence studies. The aircraft landing simulations with different free surface conditions, i.e., calm water, regular wave with different wave heights are then performed and quantitatively compared through several physical parameters, including acceleration, velocity, pressure, pitch angle and free surface deformation. It was found that the aircraft regular wave-landing process experiences several unique stages comparing with the calm-water-landing case. The results clearly confirm that wave surface can influence the aircraft landing performance to a great extent. The fundamental mechanism is found to be that the wave surface slope and wave particle velocity remarkably change the impacting position and effective impacting velocity of the aircraft.
Article
The objective of this program was to (1) review and analyze worldwide transport accident data relative to water impacts and ditching performance, (2) compare the results of this study with current FAA requirements to determine their adequacy/relevancy, and (3) conduct a survey of major worldwide airports to determine their proximity to water. The data were analyzed with respect to the airplanes' structural integrity, breakup patterns, subsystem performance, cabin integrity, and airline procedures that were or could be contributors to injuries and fatalities. A summary of the relevant FAR's was presented to provide background for analysis of the accident data. The methods by which airframe manufacturers may certify their aircrafts' ditching behavior were presented. Because of the infrequency of unplanned water contacts and ditching (planned) occurrences, a case study approach was taken in analyzing the accident data. Eleven worldwide water impact accidents were identified between the years 1959 and 1979. Of these, only one was classified as a ditching occurrence. For the years 1980 to the present, three U.S. water related occurrences were identified with no ditching occurrences. In deep water accidents, it was found that when the flight crew had at least some degree of preparedness, trauma-caused injuries were minimized while the majority of fatalities resulted from drowning. When the impact was unexpected, however, the forces on the airplane were generally much higher, resulting in a higher proportion of injuries and fatalities caused by trauma. In shallow water incidents, usually occurring as a result of runway overruns, drowning was not as common.
The Boeing 737 technical guide. London : Tech Pilot Services
  • Brady
Brady. The Boeing 737 technical guide. London : Tech Pilot Services, 2007.
CNN interactive. Ethiopian airline crash kills at least 50
  • G W Brune
  • J H Mcmasters
G. W. Brune and J. H. McMasters. Applied Computational Aerodynamics, volume 125 of Progress in Astronautics and Aeronautics, chapter 10 "Computational Aerodynamics Applied to High Lift Systems", pages 389-433. American Institute of Aeronautics and Astronautics (AIAA), Washington D.C., 1990. CNN interactive. Ethiopian airline crash kills at least 50. http://edition.cnn.com/WORLD/9611/23/hijack.ethiopia.pm/index. html Last checked: 01.05.2009, Nov. 1996a.
ET-AIZ) Aircraft Accident In The Federal Islamic Republic Of The Comoros
Ethiopian Civil Aviation Authority -Flight Safety Department. Aircraft Accident Investigation Report Ethiopian Airlines B767 (ET-AIZ) Aircraft Accident In The Federal Islamic Republic Of The Comoros, In The Indian Ocean On November 23, 1996. Addis Ababa, Ethiopia, May 1998.
Federal Aviation Regulation Part 25 Airworthiness Standards: Transport Category Airplanes
  • A Evans
A. Evans. Accident to Boeing 707-351C ST-APY near Mwanza Airport on 3rd February 2000. Flight record replay report, Air Accident Investigation Branch (AAIB), Aug. 2000. Federal Aviation Administration (FAA). Criminal Acts Against Civil Aviation, 1996. Federal Aviation Administration (FAA). Federal Aviation Regulation Part 25 Airworthiness Standards: Transport Category Airplanes, Subpart D -Design And Construction, Emergency Provisions, Sec. 25.801 Ditching, Amendment 25-72, Aug. 1990.
International Civil Aviation Organization (ICAO). Manual of the ICAO standard atmosphere
D. E. Fink (Editor). Aircraft search for missing peruvian 727. Aviation Week & Space Technology, 133(12):42, Sept. 1990. Flight International. Altimeter discrepancies found in crashed 707. Flight International, 157(4717):7, Feb. 2000a. Flight International. Trans arabian's 707 splashes down. Flight International, 157(4715):11, Feb. 2000b. Flight International. Prepare to ditch. Flight International, 86(2892):241, Aug. 1964. A. Indra. http://www.airdisaster.com/photos/ga421/photo.shtml Last checked 01.05.2009, 2002. International Civil Aviation Organization (ICAO). Manual of the ICAO standard atmosphere : extended to 80 kilometres (262 500 feet). Montreal, 3rd edition, 1993.
Investigators probe sudden engine failure on descending Garuda 737
  • N Ionides
  • D Learmount
N. Ionides and D. Learmount. Investigators probe sudden engine failure on descending Garuda 737. Flight International, 161 (4815):7, Jan. 2002.