French research and supply vessel Marion Dufrense ran aground in the morning Nov 14, 2012 in position 46 26S / 051 53E in a bay Cro-Baie du Marin, Possession island, Crozet Archipelago, Indian ocean. Vessel reportedly suffered a breach in a bow part, portside, two compartments and 5 ballast tanks are flooded.
From marine safety investigation report from BEA MER :
GROUNDING OF THE OCEANOGRAPHIC VESSEL MARION DUFRESNE
on 14 November 2012 in the Western Approaches of Ile de la Possession (Crozet Archipelago)
Report published on : November 2013
As part of her third mission of the year for the TAAF, MARION DUFRESNE, who sailed on 7 November 2012 from La Réunion with 97 passengers and 48 crew members on board, arrived during the 13 November night in sight of Île de La Possession, in Crozet Archipelago.
She cast anchor at 0.45 mile in the North of Pointe Basse at 5.50 am.
The unloading of the equipment for the scientists’ shelter had been achieved around 8.00 am.
The vessel sailed bound to another shelter in front of which she would cast anchor for a similar operation.
Baie du Lapérouse, located in the south-west of the Island, is a dozen miles away.
The course shaped was rounding the western end of the Island.
A helmsman was steering.
At 8.44 am, in the west of Pointe des Moines, the master handed over the manoeuvre to the OOW.
A few minutes later, when the vessel was closing the waypoint between the 208° and 125° courses plotted on the chart (WP28), the OOW began to alter course.
Considering it to be too early, the master took over the manoeuvre back and ordered to set the helm midships then a few moments later, hard-a-starboard.
An impact was then felt the vessel took on a starboard list.
The speed of the vessel suddenly decreased from 11 to 6 knots.
The master reduced speed to slow ahead.
He requested to close the watertight doors and to activate immediately the damage control procedures.
Soon before 9.00 am, several compartments were reported (remote water level sensors and alarm control panel) flooded by seawater, the vessel was normally floating and making her way slowly ahead.
The master decided to proceed to his final destination, the Baie du Marin, located at the eastern end of the Island opposite to Port Alfred, where is the main research station of Crozet archipelago, well protected from the westerly swell.
He cast anchor there around 11.00 am.
Until 17 November, all the passengers and the half of the crew members had been taken ashore on the Island, in very rough weather conditions due to a succession of strong low pressures.
On 23 November, the French cable vessel LÉON THÉVENIN, chartered for assistance, arrived off Baie du Marin and sent divers ashore.
On 24 November, they carried on investigations of the hull which led BV to give, on the same day, its agreement for a single voyage.
A provisional navigation license had been issued for the voyage to South Africa by the vessel safety centre of La Réunion in the early evening.
MARION DUFRESNE sailed immediately, escorted by the tug CORAL SEA FOS sent in the area.
On 3 December, around 5.00 am, the vessel arrived on roads at Durban.
After some controls (particularly for pollution risks) by the harbour authorities, she was allowed to enter the port where she came alongside in the early afternoon.
The technical stop had lasted almost two months.
extract from SHOM chart 6497
MARION DUFRESNE is owned by GIE MD II (99% TAAF and 1% CMA).
The vessel is bare boat chartered by CMA CGM, which ensures the technical management and equipment of the ship.
The ship management is entrusted to CMA Ships, a CMA CGM subsidiary.
TAAF charters the vessel under a twenty year time charter contract until 2015.
The Institut Polaire Français Paul Émile Victor (IPEV) charters the vessel to TAAF 217 days per year for oceanographic surveys in all the oceans, off-ice.
All the passengers are managed by the TAAF during logistical voyages towards Austral Islands and aboard under the master’s responsibility.
The charterer (TAAF) is represented aboard by the person in charge of Operations and Austral Expeditions (OPAE).
He supervises all the persons aboard who are not part of the crew: scientists, persons on mission, in transit, or passengers.
After GALLIENI then the first MARION DUFRESNE, her predecessors, the second MARION DUFRESNE, has been designed for the transportation of personnel and equipment to the territories (TAAF public service mission, 120 days per year) and, in addition, to participate in various oceanographic surveys (IPEV).
She has been specially built in 1995 to perform this multi-purpose task.
Notice particularly, that the subdivision of the double bottoms and the bilge pumping system had been specially designed to ensure the vessel’s buoyancy, taking into account the operation areas, unsurveyed or poorly surveyed, and submitted to inclement weather conditions.
Studies are on-going to prepare a major technical stop in order to extend her career for fifteen years.
The mission (OP3-2012) in which the vessel was involved at the time of the accident was a TAAF mission, one of the four undertaken each year by the TAAF in Crozet Archipelago, in March, August, November and December.
In the TAAF area, the chart coverage and the navigational documents are based on a limited knowledge of the local hydrography.
MARION DUFRESNE has been operated intensively in rough sea conditions since her commissioning in 1995 as she has been at sea or mooring more than 93 percent of the time with an average number of passengers of 56 per day.
She is assessed by her operators and users as a reliable and safe vessel.
She has two work launches (tenders) used for passengers disembarkation or embarkation, but also for towing equipment on « portières » (cf.
glossary) and for hydrographic surveys.
They have to be operated in sight and in VHF radio range from the vessel, for navigations restricted to sheltered waters.
During TAAF support missions, she is fitted with a helicopter.
Several scientific laboratories are aboard MARION DUFRESNE, who is fitted with a multibeam sounder and a giant corer (60 m).
SEQUENCE OF EVENTS
Local times (UTC + 4)
On Wednesday 7 November 2012
At 5.12 pm, MARION DUFRESNE sailed from Le Port (La Réunion) bound for Île de La Possession (Crozet Archipelago) with 48 crew members and 97 passengers (TAAF and IPEV members as well as 9 fare-paying passengers).
On Wednesday 14 November 2012
The morning weather conditions were as follows (source: SITREP):
Cloudy / westerly wind force 4 / slight sea state / small swell / good visibility.
At 5.00 am, the master came up to the bridge.
At 5.50 am, the vessel was at anchor at 0.45 mile in the North of Pointe Basse (Île de La Possession).
Equipment for surveys and scientific works had been carried by helicopter to a shelter ashore.
At 7.40 am, the « lieutenant Océano » came up to the bridge to take over the watch from the prevention officer who was in charge of the watch from 4.00 am to 8.00 am.
She informed him of the situation (vessel still at anchor) and showed him the successive courses to be followed towards the Baie du Lapérouse anchorage at the south-western end of Île de la Possession for a second operation scheduled around noon.
At 8.00 am, the « lieutenant Océano » took over the watch, the prevention officer remained on the bridge.
At 8.08 am, the operations were achieved and the helicopter was secured in its shed.
At 8.16 am, the anchor was aweigh and the vessel got underway under the master’s con towards Baie du Lapérouse.
The helm was under manual control.
At 8.36 am, the OPEA requested the passengers to leave the left wing of the bridge to facilitate the work of the crew, the passengers complied.
At 8.41 am, a new group arrived and settled.
The OPEA requested them to move away or to go up to the upper bridge.
The master stayed on the bridge, assisted by the OOW, the helmsman was steering.
At 8.44 am, the vessel rounded Pointe des Moines at 0.55 mile in the west (radar fix).
The master handed over the responsibility of the watch to the OOW who took it up, he drew his attention on the breakers and went on the bridge port wing.
The helmsman kept on steering.
The vessel was heading 212° at 11 knots.
The course laid off was 208°, the radar fix was on the inshore side of the laid off course, the course was altered to come back on the initial track.
At 8.52, the OOW made a radar fix and plotted it on the chart.
The position plotted showed that the vessel was at 0.2 mile on the inshore side of the laid off track and at 1 mile from the WP n° 28 (in the west of the breakers) marking the next course alteration point.
At 8.53 am, the OOW ordered the helmsman: « port ten ».
As the vessel was veering quickly on port, the OOW ordered: « Steady as she goes » (i.e.
steer on the course of the vessel at the time of the order).
The helmsman « met the helm » quickly in order to keep on at the course which was then 180°.
At 8.55 am, assessing to have the breakers abeam, the OOW would have ordered (he does not remember to have given this order) to the helmsman to turn to the heading 125°.
The helmsman came gently on port.
Seeing the vessel veering and understanding that the course alteration was premature, the master came back to the centre of the bridge.
He observed that the vessel was frankly on the inshore side of the laid off track.
At 8.56 am, the master ordered then the helmsman to set the helm midships, which he did.
A few seconds later, he ordered to put the helm hard-a-starboard and at the same moment an important impact was felt on the bow.
The vessel heeled violently on starboard.
The speed decreased from 11 to 6 knots.
The master slowed down to « slow ahead ».
At 8.57 am / 8.58 am, as ordered by the master to the OPEA, the passengers left the bridge silently and mustered in the scientific control room.
At 8.59 am, the vessel was normally « afloat » and was making her way slowly.
The initial emergency measures had been immediately implemented:
- The watertight doors had been closed;
- The bilge pumps had been set on;
- The passengers still on the bridge had been evacuated;
- Patrols had been undertaken in the engine room as well as a check of the bow
capacities in order to identify damages.
The master decided to cancel the call scheduled at Baie du Lapérouse anchorage
and to proceed to Baie du Marin, which is located in the south-eastern end of the island, opposite to Port Alfred (Archipelago main base) and relatively sheltered from severe weather coming frequently from the west in this area.
From 9.00 am to 10.00 am, the chief engineer, who was on the bridge, ordered to carry on the pumping of the flooded compartments while the crew was investigating.
At 9.02 am, the watertight doors were remotely closed.
At 9.05, the master triggered the VDR data backup, and headed to Baie du Marin,
sheltered from the north-westerly wind to cast anchor safely.
From 8.57 am to 10.14 am, a succession of numerous alarms occurred:
- 8.57 am / NH (high level) cofferdam;
- 8.58 am / NH echo sounder room;
- 9.03 am / NH bilge 1 port well;
- 9.21 am / NH bilge 1 starboard well;
- 9.30 am / NH bow thruster room;
- 9.39 am / echo sounder room flooded;
- 9.48 am / NH fresh water cross-pipe ;
- 10.14 am / bow thruster room flooded.
At 9.45 am, rounded « Cap Gallieni » at 8 cables in the south,
At 10.02 am, first call to La Réunion MRCC,
At 10.45 am, the vessel was anchored with 5 shackles in the water, in Baie du Marin, at 0.5 mille in the south of a position defined by the crossing of the alignments.
At this time, the assessment of most of the damages was achieved.
At 11.33 am, first call to the owner on the emergency line.
From 2.40 pm to 2.50 pm, 16 passengers had been landed in 4 helicopter rotations.
An ongoing monitoring of the flooded fore compartments had been established through the permanent presence of personnel on a four shift basis, in constant liaison by walkie- talkie with the bridge.
During the 14 November, the wind increased to west-south-westerly 8, the master paid out two more shackles, the holding was good.
On Thursday 15 November 2012
At 7.30 am, north-westerly near gale, a visual draught survey had been done at anchor from the Zodiac: 5.60 m F; 6.70 m A.
At the same time probes showed that except for GO4, all the ballast tanks affected by the event were full.
At 10.00 am, the bow thruster room was bailed out and made temporarily watertight.
At 11.00 am, the water level in the echo – sounder room began to decrease.
From 4.00 pm to 4.20 pm, 20 passengers had been landed in 4 helicopter rotations.
In the evening winds freshened to Force 8.
On Friday 16 November 2012
The weather conditions improved in the early morning.
Equipment unloading operations:
- from 6.40 am to 7.49 am, with the launch and the « portière »;
- from 9.15 am to 0.25 pm, with the helicopter;
- from 3.20 pm to 5.39 pm, landing of the 61 passengers still aboard.
In the evening, north-westerly gale, the anchor was holding good and the drainage of the flooded compartments was carried on (bow thruster room and echo – sounder room).
On Saturday 17 November 2012
The wind was falling, veering south-south-easterly.
A storm was forecast for the next day.
From 3.00 pm to 3.35 pm, landing of 20 crew members.
Only volunteers (23 persons) stayed on board.
The drainage of the flooded compartments was carried on, the breaches were made watertight with cofferdams.
On Sunday 18 November 2012
West-north-westerly gale becoming quickly a violent storm veering southerly, with 75 knot gusts of wind.
Pictures taken by the master on 18 November in Baie du Marin
Anemometer display on the same day
On Monday 19 November 2012
The weather conditions fell to west-south-westerly near gale, with 40 knots gusts of wind.
The anchor had dragged on a hundred metres (later on, it had been observed that one of the links of the cable had been distorted, it is probable that it happened at this time).
The wind fell during the night and veered north-westerly.
On Wednesday 21 November 2012
At 1.50 pm, the Zodiac was launched for a survey of the draughts, which remained
On Friday 23 November 2012
Island supply operations by helicopter.
On Saturday 24 November 2012
At 2.24 am, arrival of the cable ship LÉON THÉVENIN.
MARION DUFRESNE left her
mooring spot to the latter.
At 5.44 am, CORAL SEA FOS arrived in the area.
From 6.00 am to 7.55 am, the passengers were transferred from the base on Ile de La Possession on board LÉON THÉVENIN with the Zodiac and the launch.
From 9.20 am to 1.40 pm, diving operations under the hull in order to assess the damages.
The pictures taken had been transmitted to BV which maintained her class.
At 8.25 pm, the CSN of La Réunion issued a navigation licence limited to a single voyage to a shipyard for repairs.
MARION DUFRESNE began to heave up anchor.
At 8.35 pm, the anchor broke out, the vessel was under way bound to Durban or Mauritius.
At 9.00 pm, « Full away » signal.
The vessel was escorted by CORAL SEA FOS.
0n Monday 26 November 2012
As of now, the destination was set as Durban.
From Sunday 25 November to Sunday 2 December 2012
At the beginning of the week, succession of westerly to north-easterly gales, sea rough to moderate.
The vessel was labouring due to pitching and rolling which was causing fatigue to the crew.
From Wednesday, weather permitting, 160 metric tons of fuel had been transferred in order to empty, as far as possible, the damaged tanks.
On Monday 3 December 2012
At 7.30 am, End of « Full away ».
From 08h30 am to 10h00 am, inspection on roads of the vessel by the South-Africa authorities.
At 10.00 am, underway then waiting.
At noon, Authorisation to enter the port of Durban granted.
At 1.42 pm, finished with engines, starboard alongside at mole 1 berth 102.
At this time, the vessel had been « weighted »: her displacement was 9 100 metric tons instead of 7 800 metric tons corresponding to the displacement observed at the time of the previous dry-docking.
The technical stop for the repair of the damaged hull had been 2 months long.
The method selected for this analysis is the method usually employed by BEAmer for all its investigations, in compliance with the « Code for the Investigation of Marine Casualties and Accidents » laid out in Resolution MSC 255(84) adopted by the International Maritime Organization (IMO).
The factors involved have been classed in the following categories:
natural factors ; material factors ; human factor ; other factors.
In each of these categories, BEAmer investigators have listed the possible factors and tried to qualify them relatively to their characters:
certain, probable, hypothetical ; causal or underlying ; circumstantial, inherent ; aggravating ;
with the aim to reject, after examination, factors with no influence on the course of events and to retain only those that could, with a good probability, have a real influence on the course of facts.
The investigators are aware that maybe they have not given an answer to all the issues raised by this accident.
Their aim remains to avoid other accident of the same type; they have privileged with no a priori an inductive analysis of the factors which have a significant risk of recurrence due to their inherent character.
It had been impossible for technical reasons to extract data from the VDR although the device had been analysed by its manufacturer after the event.
The factual elements on which is based the analysis of this event are therefore the master’s sea protest, the documents collected on board the vessel, the pictures of the damages and the testimony of the crew members and of the available passengers.
picture CMA-CGM 2009
When the vessel touched the bottom, by daylight, the weather and sea conditions were good as well as the visibility (cf.
sequence of events).
On 14 November 2012, the tidal range was 107, the low-water took place at 7.07 am (UTC + 5) with a 0.05 m height of water and the high water at 1.10 pm with a 0.30 m height of water.
Neither the range of tide nor the tidal stream could therefore have contributed to the
The vessel had been able to free herself quickly from the reef.
No natural factor had any direct influence on this event.
The vessel is fitted with the regulatory navigation equipment.
Nevertheless, several factors related to the navigation equipment and to nautical documents used when touching the ground have been identified:
The navigational equipment
The GPS navigation satellite receivers:
The bridge is fitted with 7 GPS receivers connected to the various navigation or communication devices, 3 of which are directly used for navigation.
The reference ellipsoid is always WGS 84.
The GPS displays used for navigation are arranged as follow:
- The DGPS Furuno navigator GPS 90 is fitted on the navigation centre console;
- The navigation GPS Trimble NT 100 and NT 200 are fitted on both sides of the chart table and give the position to the plotter and to the 3 and 10 cm radars.
The wheelhouse console located on starboard with the plotter on the left and the ECS on the right.
On the « Nautoplot » plotter, the SHOM 6497 paper chart had been set at 5.00 am on 14 November and the changes made to match the plotter and the GPS geodesic coordinate.
The proper use of this plotter supposes that the chart has been perfectly positioned and that its positioning is checked on a regular basis.
Picture of the Nautoplot plotter taken by a passenger.
It is on the top of this chart that are mentioned the changes to the geodesic system, which are used for its positioning.
It appears on the picture below that at the time of the grounding, the positioning of the plotter was taking account of the last change done.
Corrections « done aboard » written on the paper chart SHOM 6497.
The second one corresponds to an observation done by the crew before the exact values had been received from the SHOM.
However it had not been taken into account for the use of the ECS.
The ECS (TRANSAS navigation system), which display is located on the right of the plotter, is connected to the gyro compass in use, to the Doppler log and to a GPS receiver.
On a label placarded on the left hand side of the control keyboard, it is specified that the ECS is not an approved navigation system.
This has been reminded by a circular of the company technical service.
The type of charts and their updating
The charts which are in the Transas system memory are mainly Brittish charts.
The changes are sent aboard on a CD by Transas via a Belgian provider.
The French chart changes (and particullarly chart 6497) are unlikely to be taken quickly into account.
On the left, an extract of SHOM chart and
on the right an extract of the chart used by the ECS TRANSAS system.
On chart SHOM 6497, the 100 m line of soundings which is located in the west of cap de l’Héroïne is a broken line.
This means that the position of this isobath is unaccurate.
It is observed on the ECS display that, although the reference mentionned is chart SHOM 6497, this 100 m line is a continuous line (see appendix page 48).
The chart used by the ECS is therefore slighly different from the reference chart as at least one information is found lost.
This unaccuracy is an underlying factor of this event.
Curently, for your information, the IHO has defined the ENC (Electronic Navigation Chart), a vector electronic chart produced by an accredited hydrographic service conforming to S57 norm, and the RNC (Raster Nautical Chart) produced by an accredited hydrographic service conforming to the norm - The UKHO producing a range of charts in this format, baptised « ARCS » (Admiralty Raster Chart Service).
There is at present no medium or large scale ENC or RNC covering Crozet archipelago.
These official charts, used with an ECDIS are the only ones to comply with the requirements of regulation V-19 § 2.4 of the SOLAS Convention about the chart carriage requirements, RNC are permitted exclusively in areas where ENC of an appropriate scale are not available.
The other charts, named ECS (Electronic Chart System), are not admitted by the SOLAS Convention.
MARION DUFRESNE’s ECS , even fitted with official electronic charts, cannot thus meet the chart carriage requirements as defined in the SOLAS Convention.
That was, as it happens, clearly stated on the recommendation displayed on the device.
The navigation equipment (radars, sounders, plotter, etc.) worked correctly.
On the other hand, the ECS is not approved for navigation (see above).
A navigational equipment malfunction cannot be thus retained as a causal factor of this accident.
Nevertheless, the significant error (which will be assessed later) observed on the ECS display contributed to the grounding.
Taking into account the too much exclusive use of this equipment for the navigation, this error will be studied as a human factor.
chart 6497 Ile de la Possession, île de l'Est (ed n°2 23/11/1989, scale 1:75,000)
The nautical publications : The Chart
The chart in use was chart SHOM F 6497 « Archipel des Crozet – Groupe de l’Est ÎLE DE LA POSSESSION – ÎLE DE L’EST ».
Scale 1: 75,000.
This chart had been issued complying with the former norms according to geodesic system IGN 62.
A magenta printed advice on the top right mentions: « The positions referenced to the WGS geodesic system got from a satellite positioning system (for example from a GPS receiver) must be corrected to agree with this chart.
See nota: Satellite positioning.
This nota appears on the top left of the chart in the form of an affixed chart appendix.
In 2012 two changes had been notified by a notice to mariners and made with a chart appendix:
The first, which date back to July:
X 12 27 106.
- Geodesic systems.
Chart appendix (card, 12-69 SHOM/DOPS/MIP).
— Chart 6497 (10)
Affix the chart appendix F.20 46 11.20 S – 51 42.00 E (central position)
The chart appendix F 20 corresponding to change 27106 (i.e.
change number 106 issued during week 27 - from 1 to 8 July 2012 - ) is as follows : « Satellite positioning: The positions referenced to the WGS 84 geodesic system got from a satellite positioning system must be corrected of 0.03’ southwards and of 0.02’ eastwards to agree with this chart ».
The second, which dates from the beginning of September:
X 12 37 107.
- Chart appendix.
Geodesic systems (card, 12-69 SHOM/DOPS/MIP).
— Chart 6497 (11)
Affix the chart appendix F.30 46 11.20 S – 51 42.00 E (central position)
The chart appendix F 30 corresponding to change 37107 (i.e. change number 107 issued during week 37 - from 10 to 16 September 2012 - ) is as follows:
« Satellite positioning: The positions referenced to the WGS 84 geodesic system got from a satellite positioning system must be corrected of 0.08 ’northwards and of 0.37’ eastwards to agree with this chart.
These changes (27106 from July 2012 and 37107 from September 2012) do not add up, the second includes the first.
At the time of the accident, two changes (see page 23) were affixed on the chart, one (12 27 106) received from SHOM and the other coming from the crew’s observations who doubled the error given by the first.
This paper chart was located on the Nautoplot plotter used and positioned in accordance with the last plotted observation.
The OOW stated that he did not take it into account, at the moment of the accident as he did not use the plotter, but the ECS.
Even though the observation had been plotted on the paper chart (used on the plotter), the change, issued by the SHOM about 2 months before the accident, had not been used for navigation by MARION DUFRESNE’s officers (information not yet received on board).
The error between the Radar fix and the GPS position plotted at 8.52 am represents the value of the correction of the chart appendix F 30.
The difference of geodesic reference between chart 6497 and the navigational equipment, which had not been taken into account by the OOW, as he used the ECS and not the up-to-date chart, constitutes an underlying material factor of this event.
Sailing directions Volume L9 Edition 2012 - Page 432 INDIAN OCEAN ISLANDS (Southern Part).
The sailing directions
In the sailing directions L9 « INDIAN OCEAN ISLANDS (Southern part) », one can read in the 2001 edition in the chapter dedicated to Ile de la Possession in Crozet archipelago page 329 § 188.8.131.52 (North-West and South-West coasts / 07) : «Cap de l’Héroïne, (...) Breakers located at 2 M in the SW of the cape, probably indicate hazardous rocks.
This sentence has not been modified by a notice to mariners.
In the 2012 edition of these sailing directions, published at the end of September 2012, which had not yet been received aboard, the following information is written in the chapter dedicated to the same island, at page 432 § 184.108.40.206 (North-West and South-West coasts / 013): « At about 2 M in the SW of Cap de l’Héroïne, western end of the island, breakers probably indicate hazardous rocks.
Sailing directions Volume L9 Edition 2001 - Page 329 INDIAN OCEAN ISLANDS (Southern Part).
One notices notably that the word « about » has been added concerning the position of the breakers relatively to Cap de l’Héroïne, which takes into account the changes transmitted by the Notice to Mariners number X 12 27 106 and X 12 37 107 reminded in the previous subparagraph concerning the geodesic system.
The reading of this extract from the sailing directions stresses, if proofs were needed, the need that the track laid takes into account the inaccuracy of the position of hazardous rocks and for the OOW to follow the laid track.
The position of the breakers
The « Brisants » (breakers) which appear on chart 6497 and which are mentioned by the sailing directions L9 were not visible on the day of the grounding due to fair wind and sea conditions.
Years ago, they had been reported to the SHOM by the first MARION DUFRESNE who, after GALLIENI was then the main source of nautical information in the TAAF and was thus upstream from SHOM.
Nowadays, the amount of information provided to SHOM by MARION DUFRESNE is significantly less.
The few items collected by the SHOM come, actually, from vessels of the TAAF and those of Marine nationale.
For your information, the instruction from the prime minister about the collection and the broadcast of the nautical information n°228/SGMer issued on 03 May 2002 designates the captains of military vessels and the masters of the vessels among the authorities responsible for the search, the collection and the transmission of information which could alter nautical publications.
Following this maritime casualty, the SHOM, concerned by the safety of navigation in the areas where the bottom is not fully charted, has ordered satellite pictures of this coast of Île de La Possession.
The processing of these pictures had been complex, because the area is frequently cloudy and the submerged rocks are difficult to identify.
The picture above, got on 26 March 2013, overlapped with the marine chart 6497, proves that the breakers are located on the marine chart at 443 m in the 049° of the detected reef (this picture had been provided to BEAmer by the SHOM).
This discovery was the object of change 19158 (issued during week 19, i.e. from 5 to 11 May 2013) to chart 6497:
13 19 158.
Île de La Possession.
In the SW of Cap de l'Héroïne.
(SHOM, FE N°24 DOPS/MIP/GEO/NP issued on 11/04/2013).
— Chart 6497 (12) draw « Brisants » (breakers) (a) 46 25,18 S 51 36,96 E
Delete « Brisants » (breakers) in the vicinity of (a) in the NE
MARION DUFRESNE could not, in November 2012, be informed of the true position of these breakers.
They have been re located in the vicinity of the broken 100 metres line of soundings, just on its inshore side.
The inaccurate position of the breakers on chart 6497, and though on the ECS display, constitutes also an underlying material factor.
Humans factors : Laying the track on the map
The passage plan had been established in January 2009 and had been the reference to lay the courses for this voyage.
This laid off track left little room for error given the uncertainties raised by the sailing directions (cf.
appendix C) in this unsurveyed and unmarked area.
Given the error observed by the SHOM between the horizontal datums which has been published by a notice to mariners, the track laid off according to the passage plan established in January 2009, had a reduced margin of safety at this place.
The laid off track constitutes thus an underlying factor.
The navigation itself
Courses were laid off by the navigation officer under the control of the master to round the island to the west, at the edge of the 100 metres line of soundings.
The major identified and borne hazards were the breakers of Pointe Basse located at 0.4 mile of the anchoring point, the Île de La Roche Percée which was rounded at 0.4 mile and the breakers located in the south-west of Cap de l’Héroïne that the course laid off rounded at 0.8 mile to the west.
The ECS is based on chart SHOM F 6947 and on the display, the vessel was, soon after 8.55 am, close to WP 28, in the vicinity of the label « Brisants » (breakers).
This picture is the result of the overlapping of chart 6497 processed by the SHOM, data from satellite imagery which shows the real position of the shoal and the chart used by the ECS.
On the picture:
the yellow track represents the course laid off according to the passage plan; the black track represents the plot of the positions recorded by the ECS;
the red track represents the actual course on the ground reconstructed by BEAmer given the real position of the collided reef, the grounding sequence of events and the damages observed on the hull of the vessel (port bow).
The white frame is zoomed on next page.
The OOW followed the black track while he was actually on the red one.
He was therefore at 0.38 mile in the 049° of the position he thought to be his (i.e.
0.24 mile more north and 0.43 mile more east).
The processing of the information « Brisants » (breakers) on chart 6497 supposed that the two following correction items had been taken into account:
- The first, « 0.03’ of latitude towards the south and 0.02’ of longitude towards the east ».
- The second, « 0.08’ of latitude towards the north and 0.37’ of longitude towards the east » which cancels and replaces the first.
They were available before the accident on the website of the SHOM.
These correction items were not taken into account for the navigation with the ECS (non approved equipment for navigation).
On the chart above, the changes are indicated by the points A1 and A2.
- The third element lies in the choice of the ECS for the navigation :
The impact of the vessel’s hull on the north-western part of the reef (real position SHOM) leads to a difference with the position indicated by the ECS of « 0.17’ of latitude towards the north and 0.02’ of longitude towards the east ».
This error is thus attributable to the ECS.
On the chart below, the grounding position is indicated by point A3.
The error between point B (« Brisants » (breakers) indicated on chart 6497 before change 13.19.158) and point A3 is the result of the repositioning of the reef done by the SHOM which put the hazard closer to the 100m line of soundings.
This information, later than the accident, was not available at the time of the event.
The failure to take into account the difference of horizontal datum for the navigation with the ECS at this position, the poor management of the inaccuracy of the positioning of the breakers and the use of the ECS as the unique navigation aid are elements the addition of which is a causal factor of the grounding.
It has not been possible for BEAmer to analyse to what extend the depth indications got from the echo sounder had been used for navigation before touching the ground.
It is certain that because of the vicinity of broken isobaths, the use of the echo sounder would have given an alert for the vicinity of the shoal.
The course alterations
From the previous way point, WP27, taken on the inshore side of the laid off track, the vessel had always been sailing on the inshore side of this track, at about 0.4 mile.
At 8.44 am, the OOW had taken over the effective responsibility of the watch a few minutes before the accident, but he had been on the bridge for more than one hour and he laid off himself the radar fixes on the chart since getting underway.
When the master handed over the con to the OOW, he reminded him the presence, on the chart, of breakers to round (this had been confirmed by both of them).
Therefore he had these well in mind and he knew that he should alter course to 125° in the west of these breakers in order to head to Baie du Lapérouse anchorage.
The cross track error was then similar to the error between the equipment horizontal datum WGS 84 and the chart horizontal datum IGN 62, i.e. almost 0.4 mile.
Although GPS fixes positioned MARION DUFRESNE on the inshore side of the laid off track, the OOW anticipated the course altering by cutting the corner between the current heading 213° and the next 125°.
It appears that at this time, the OOW was confident in the position displayed on the ECS to the detriment of the positions that he laid off himself on the chart from radar bearings and distances.
This ECS picture had been recorded at 8.52 am, a moment before the vessel altered course to 180°.
The heading was then 213°.
The reference displayed on the device is chart 6497, but the line of soundings appears as a continuous line.
The anticipated course alterations compared to the laid off course, approximately sailed along, constitute an aggravating factor of the previously identified factors.
However, the fact that the master stopped the turning, when he realized what manoeuvre was on going, saved the vessel.
This picture of the display had been taken a few days after touching the ground
while the vessel was approaching Durban for repairs.
Other factors: The choice of the navigational documents by the company
The company uses mainly the « Admiralty charts and publications » whose updating is transmitted by its provider through the Internet.
The chart of Crozet archipelago is not part of the « Admiralty portfolio ».
The only existing chart is the SHOM chart whose horizontal datum is not the WGS 84.
The changes to the SHOM charts are downloaded by the competent department (NAVCOM) of the company on the SHOM website and transmitted to the vessel on a compact disc.
The SHOM last change (n° 12 37 107 broadcast and available on the Internet on 15 September 2012), concerning the latitude and longitude errors related to GPS coordinates (which happened to be close to the indications laid off on the chart on 31 August 2012 (0.1’ N and 0.3’ W)), had not been downloaded by the vessel at the time of the grounding.
This constitutes an underlying factor.
In this respect, it should be noted that the vessel does not have any Internet access for her administration and nautical management, and that the chart appendix F30 had not been included in the corresponding sending by the « Navcom » department of the company.
Reconstruction of the courses on paper chart 6497
The ship-owner installed in the month following the event:
- a new VDR on board;
- a new process of assessment of all the sailors joining the vessel and at the time of recruitment.
- In the frame of a training and familiarization project :
A course « Management and psychological risks » for masters and senior officers;
A specific training for entrants in the company;
An improvement of simulator training.
- In terms of crisis management, the training of persons in a manager position
aboard MARION DUFRESNE (sailors or not).
- A crisis management plan in the Terres Australes.
The ship-owner has reminded, in the master’s permanent orders, the latitude given to the OOW, to request people not involved in navigation to leave the bridge whenever it is necessary.
According to circular MSC 1/1222, the VDR is subjected to an annual visit (the last one took place on 18 May 2012 and the test proved conclusive).
On 14 November 2012, nearly one hour after getting underway from the anchorage at Pointe Basse in the north-west of Île de la Possession, bound to the next anchorage, at about two hours sailing, the master handed over the watch to the OOW.
Assessing that his position allowed it, the OOW altered course, cutting the corner, from heading 208° to heading 125°.
This premature decision had been taken in the following context:
- The vessel was sailing along a track on the inshore side of the laid off track;
- The position of the shoal touched by the vessel was not known accurately;
- The last change concerning the horizontal datum of the chart in use had not been taken into account in the ECS;
- The too exclusive use of the ECS.
The last minute manoeuvre initiated by the master did not prevent from touching the ground but saved the vessel.
After the emergency measures and the damage assessment achieved, the vessel sailed to Durban for repairs, after the passengers had been landed.
This event could have had worst consequences if the crisis management on board and ashore had not been perfectly controlled, especially as, in this remote area, the weather conditions are often severe.
The BEAmer reminds:
To ship-owners, masters and navigation officers:
1- 2013-R-053: navigation equipment not approved by the IMO should be used with prudence, and the positions have to be systematically cross-checked by other approved systems (radar), particularly in confined waters or close to the shore.
To persons in charge of maritime training:
2- 2013-R-054: to warn, even more, cadets of the risks inherent to the use of navigational electronic aids (particularly the GPS) without cross-checking with visual and radar observations.
The BEAmer recommends:
To the CMA CGM company and more generally to ship-owners:
3- 2013-R-055: to make sure that the charts, particularly electronic charts and the sailing directions are kept up-to-date, underlining that the use of such a documentation should be done with the concern to select the best available source in the geographical area of operation.
To the SHOM:
4- 2013-R-056: to republish the chart SHOM 6497 with the horizontal datum WGS84.
To the CMA CGM company and owners of vessels carrying passengers:
5- 2013-R-057: to develop written procedures in order to preserve, on the bridge, the quality of the lookout and the necessary concentration for the practice of navigation, particularly during manoeuvres and in hazardous navigation areas.
To the TAAF Prefect:
6- 2013-R-058: to coordinate the requirements for nautical information and for buoying in the TAAF in cooperation with the maritime players interested in navigation in these areas.