The M/T 'Erika' tanker incident outside Brittany the 11 and 12 December 1999 has provoked questions. The Erika broke apart on 12 December 1999, while underway at sea and spilled >10 000 tons of heavy fuel oil. Why? A very stupid answer was given 21 February 2005!
The M/T 'Erika' was damaged, unseaworthy, prior to the last voyage!
'Experts' (no names available) of the French judge Dominique de Talancé have 2001-2004 concluded that the accident was associated with cracks/fractures in the main deck, cracks/fractures that were visible during the vetting inspection of the surveyors of the charteres, Total, just prior to the voyage! And because of this fact (sic) regarding the sub-standard, alleged 'garbage can' tanker, Total should not have chartered the tanker. But, because Total or its vetting surveyor was ignorant of the importance of an alleged crack in the main deck, Total is one way or another implicated or responsible of the oil spill! Decision by the Paris Appeal Court 21 February, 2005! Strange story! And a final judgement was made 16 January 2008. Read the background (or jump to the End 2005/2008):
The official investigations (see below - totally ignored by the French courts) have explained why the M/T 'Erika' broke apart but the reasons are not convincing. Another cause could have been a fracture in the single hull side shell internal structure, between old and new repaired steel, caused by fatigue and cyclic, heavy wave loads on the side on the morning of 11 December. The tank structure was then locally damaged - see picture right of a tanker that was saved in time! Some plates were buckled and/or fractured. Not very serious actually. Of course it is a major damage and then you must immediately go to a safe port of refuge. Oil then started to leak and the tanker asked for assistance. However, the leaking and damaged tanker was apparently denied a port of refuge and had to stay at sea in the heavy weather for another day. The fractures therefore developed upwards to the main deck, which in turn fractured across the full beam, due to the continued bad weather. Then, on the 12 December, the fracture developed downwards through the two longitudinal bulkheads and the side shell on the opposite side of the original damage and the section modulus of the cargo tank body became zero - the only structure connecting the two parts was the bottom plate, which was ripped apart.
Thus, we may know how the M/T 'Erika' broke apart. Unfortunately we have not been told why the ship was denied a port of refuge to stop the small fracture to develop into a disaster. Because the proximate cause of the oil spill was that a safe port of refuge was refused the M/T 'Erika' !
One result of the M/T 'Erika' accident is the proposal to accelerate the phase out single hull tankers and replace them by double hull tankers. It would have been better to facilitate access to port sof refuge. Unfortunately there is no evidence that double hull tanker structure will perform better than single hull after failure. Double hull tanker structure has less redundancy than single hull and will break apart faster after an intial failure and refusal of a port of refuge. The correct political decision should be to recommend the COULOMBI EGG tanker design to replace both single and double hull.
According to the Malta Maritime Authority (MMA) October 2000 there were many concurrent causes of the M/T 'Erika' accident:
"The loss was the result of several factors acting concurrently or occurring simultaneously ... The most likely reasons for the loss were corrosion, cracking and local failure, vulnerabilities in the design of the ship, and the prevailing sea conditions. ... In 1998 the tanker underwent repairs at the Bijela shipyard in Montenegro. ... The quality of the Bijela repairs could have contributed to the initial local failure, leading to the final collapse ... The ship's managers were in attendance when these repairs were carried out, yet they failed to identify and/or address areas of significant local corrosion, nor did they monitor the repairs correctly."
According to the Malta MA there were eight causes (sic) none of them proximate; corrosion, cracking, local failure, vulnerabilities in the design, prevailing sea conditions, quality of repairs, failure of the manager to address areas of corrosion and to monitor the repairs. The repairs took place about 18 months before the accident. Normally an accident investigator should find one only proximate cause of an accident - not eight concurrent causes - but let's initially only discuss that the MMA put serious blame on the ship's managers - they failed in their duties.
To do proper repairs of an allegedly corroded tanker is not easy. Evidently severley corroded parts are renewed and the repairs stop, where the original steel is intact. The extent of the repairs is agreed between owners, ship's manager, shipyard and class. During removal of corroded parts you may discover new damaged areas and the extent of repairs is extended. Evidently you should not or cannot weld new steel to original, severely corroded steel.
The work is then supervised by the workers foreman, the shipyard's Quality Assurance team, the ship's manager's supervisor(s) and the class. It is evident that welding of the ship's hull new steel plating to existing plates needs to be checked at every step of welding and that the preparation of works is correct at every step. Only particularly qualified welders are used for this delicate work.
Evidently it is the ship's manager's supervisor who has ultimate responsibility but the total responsibility must be shared by the yard and the class. If, which happens frequently, the managers supervisor is an ex chief engineer with limited welding experience of steel hull, and if the welding is done day and night concurrently with other work occupying the supervisor, he (or she) must rely on the shipyard and class that the steel work is done properly.
One lesson to learn is that all repairs affecting a ship's hull plating must be recorded properly.
In December 2000 the classification society RINA responded to the allegations about corrosion. RINA stated that the last special survey in 1998 did not give evidence of accelerated corrosion in the ballast tanks. Hull-thickness measurements were taken. During the 18 months that then passed until the accident RINA did not receive any information about problems and the vessel passed several oil company vetting surveys and two port state controls - no problems! RINA suggests six other causes which could have contributed to the accident to be inspected further. But maybe the fracture started in a cargo tank?
All oil from the wreck was removed during the summer of 2000. It would have been extremely helpful to find the exact cause of the accident and that the fractured parts would have been filmed at this time.
Did in fact the alleged fracture start in the interface between new/repaired and old steel, and how could it develop into a disaster, whereby the tanker split into two? Did corrosion play any part? By close-up filming of the ripped apart edges we might find the answer. Unfortunately the charters, Elf-Total-Fina, who paid for the removal of the oil from the wreck did not apparently film the edges of the two parts!
RINA 31 March 2000
The Italian classification Society RINA of the M/T 'Erika' issued already 31 March 2000 an internal report where chapter 5.1 Cause of the loss suggests:
"The results of the RINA internal technical investigation indicate that:
- the M/T 'Erika' was presumably lost because an initial crack in the low part of the hull below the water line was misjudged and mishandled allowing it to develop until the hull break-up;
- the ship was not lost because of an overall hull girder collapse but because she suffered a progressive structural failure.
The hypothesis of the initial crack below the water line, ... , explains the sequence of events as reported by the Master and entered in the Chief Officer's logbook, and it is also confirmed by the preliminary findings of ROV investigations.
From calculations, it is evident that the ship was designed to stay afloat with one tank open to the sea and that the longitudinal hull girder strength was sufficient to withstand the loads both on departure from Dunkirk, during the voyage and in the course of the casualty.
The sequence of events would not explain an overall hull girder collapse as cause of the loss having regard to the fact that the casualty did not develop rapidly but took more than 18 hours. Furthermore, the videos show that the deck was not affected by large buckling deformations until the ultimate detachment of the fore part. Only at that final stage did the deck structures collapse, under the torsional and shear forces acting on a section open to the sea on the starboard side.
The break-up of the hull, therefore, was not the cause of the loss but rather the consequence of the initial crack and the lack of response to it."
Lack of response? The Master knew he had a problem - but was denied a safe port of refuge to sort it out. Come on, RINa!
The French Investigation 2001-2004
The French appeal court judge Dominique de Talancé asked 2001 two 'expert' naval architects to determine the causes (sic) and the responsibilities of the accident. The 'experts' concluded (Figaro 5 October 2001) that the trading certificates were issued on 21 November 1998 and had not been re-newed on 21 November 1999, i.e. the tanker was trading without certificates. However, that did not cause the ship to split into two. As the Class was still valid and as the trading certificates rely on Class regarding hull strength, out-of-date trading certificates cannot have caused the accident.
So what caused the accident?
The 'experts' concluded that the tanker was not popular with the oil majors - black listed in 1993, rejected by TPS 1994-1995, accepted by Shell and BP in 1996, rejected by BP in 1997, rejected by TPS and Shell in 1998 - and that it had been detained twice - corrosion of bulkheads on 11 December 1997 and corrosion of the hull on 20 May 1998.
It would appear that the Erika then went to drydock in Montenegro May 1998 and carried out the necessary repairs under RINA supervision and then traded another 18 months before the accident. So corrosion could hardly have caused tha accident. The 'experts' do not discuss the quality of the repairs, so it may be assumed that they were in order.
The 'experts' state that the M/T 'Erika' was rejected by BP in 1999. A rejection by an oil major is not the cause of an accident, and we are not told why the tanker was rejected. And according to RINA other oil majors and two PSC's passed her.
The 'experts' blame the charterer Elf-Total-Fina for having chartered a sub-standard tanker, but it is not clear how Elf-Total-Fina according to the 'experts' should have been able to, e.g. inspect the hull and cargo tanks of the tanker and spot a weak area, where a fatigue fracture may start.
When oil majors 'vet' a tanker, the inspector checks many visible items, procedures and staff, but evidently the vetting surveyor cannot inspect all the details of the internal hull in cargo and ballast tanks or the quality of recent repairs.
Many parts of a ship remain hidden - from anybody. It takes one day to close-up inspect a cargo tank and access has to be prepared days before so proper close-up inspection of the structure in the cargo tanks is not part of any vetting. Actually a vetting surveyor cannot provide any correct opinion about the condition of the internal structure of the tanker. The vetting surveyor must rely on the shipowner's and Class' reports.
The 'experts' do not explain why the tanker split into two or how the charterer (Elf-Total-Fina) should have been able to recognize that risk or possibility a week earlier. And we still do not know why the tanker split into two. It is extremely unsatisfactory that pictures of the broken parts have not been made available.
The 'experts' do not mention that the responsability of seaworthiness (and cargo worthiness) of a tanker rests with the ship owner - not the charterer. The due diligence rests with the owners but an owner (or charterer) cannot be held responsilble for a hidden defect, which later may cause a fatigue fracture.
The 'experts' rightly blame the French (Brittany) maritime safety authority (Premar) for having misjudged the first Mayday and the request for a port of refuge. Premar did not consider the first request from the M/T 'Erika' on 11 December seriously and no emergency procedures were activated on 12 December. In that time a small fracture in the side apparently developed into the tanker splitting into two parts.
Malta, Italy and France made a big mistake during the Erika accident investigation. None demanded that the IMO resolution A.849 (20) should be followed and therefore the result was the above mess of a sloppy official investigation followed by contradictory and confusing suggestions of other causes by other parties. The correct procedure is simple - the flag state (Malta) is responsible for the investigation. Other interested States (e.g Italy, France) announce that they want to participate in the investigation and that it should be public. Finally one common report is issued. If later new facts emerge changing the official cause, these facts must also be investigated. One proximate cause of the oil spill must be declared = the refusal to allow the vessel to proceed to a port of refuge!
A French attempt of accident investigation is found at http://www.mer.equipement.gouv.fr/actual/rapports/Erika_anglais_final.PDF . It is un-dated and not complete but described well the underwater survey of the wreck.
The M/T 'Erika' accident has led to new demands year 2000 of phasing out single hull tankers and the use of Double Hull only or approved alternative design (even if the cause(s) of the M/T 'Erika' accident is (are) still not finally resolved 2001).
However, the side structure of a Double Hull tanker is generally higher stressed than single hull due to the void inside side space - no cargo pushing from inside against the side - and more prone to corrosion and fatigue. The whole side is a ballast space! A fracture in the side of a Double Hull tanker will then result in the same accident as the Erika (in similar, very unusual circumstances), i.e. the upper deck fractures first (because the section modulus is less at the deck - the fracture goes upwards), then the deck fractures and finally the longitudinal bulkheads and side shell members fracture down to the bottom. There is nothing to pull the tank body together, when the deck is fractured - double hull or conventional single hull.
In October 2000 the 1989 built Double Hull chemical tanker the Ievoli Sun sank loaded with >5 000 ton Styrol. The cause of sinking is not yet known - heavy weather? - but a contributing cause could very well have been defects of the double hull spaces/structure. Everyone (?) knows the difficulties with leaks of chemicals into the double hull spaces - how to clean in out? How to repair the leak?
The writer has inspected (on behalf of a prospective buyer) a double hull chemical tanker with a mixture of unknown chemicals in the double bottom! The particular trade of the tanker was so hectic with >ten different chemicals loaded/discharged that the crew had little time to inspect all double bottom tanks between cargoes. The result - suddenly the tanker had a mix of unknown chemicals in its double bottom? How to clean that? And where was the leak - fracture? And could the mixture have damaged the outer hull? Double hulls can be very dangerous!
However, the COULOMBI EGG design will prevent the M/T 'Erika' and the Ievoli Sun type accidents! Why? The answer is simple! When the fracture in the side of the ballast tank develops upwards to the deck, the tank body is kept together not only by the bottom structure but also by the stepped, mid-height deck structure. The section modulus with a fractured deck is not 0% but about 30% of the original due to the mid-deck structure and this is enough to prevent the fractures to proceed downwards. Evidently, any tanker is quite strong and does not suddenly break apart due to a fracture unless there are contributing circumstances, e.g. refusal by shore to permit a port of refuge, but there should be no doubt that the residual strength of the COULOMBI EGG tanker structure is greater than any other design. Also, the COULOMBI EGG tanker is always sagging when loaded, so the still water bending moment is always preventing the fracture do develop upwards.
On 28 March 2001 the 34 365 dwt bulk carrier Tern collided with the 37 000 dwt Double Hull tanker Baltic Carrier. It was a classic collision between equal size ships (used in the IMO data base to evaluate collision damage, where severe spills only occur in 25% of the cases - when the inner hull is breached - according to the statistics). It was - of course - the upper side of the Baltic Carrier that was breached and 2 800 tons of heavy fuel oil flowed out from one breached centre tank. The hull inside B/5 from the side was not damaged.
Had the Baltic Carrier been built according COULOMBI EGG principles with a B/5 collision crush zone in the upper side, it is highly probable that no cargo tank had been breached. And had a lower side tank been breached, only < 700 tons would have flowed out. Luckily the cargo on the Baltic Carrier was inflammable heavy fuel oil, which did not catch fire. Next time you may not be so lucky. The only effective severe collision protection for tankers is the COULOMBI EGG design.
Does anyone believe that Total vetting surveyors overlooked a crack/fracture in the M/T 'Erika' main deck late November/early December 1999? The Commercial Court of Dunkerque has already exonerated Total and its vetting surveyors, but the Paris Appeal court 2005 ignores common sense. Cargo worthyness of a tanker is now the responsibility of some vetting surveyors. The ship owner, the master, the port installation staff, the various administrations, Port State Control, the Class society - none is responsible except the vetting surveyor?And the vetting surveyor noted a crack in the main deck of the M/T 'Erika' - and informed Total ... and said it was OK!?! You wonder where this vetting report is? But they did not tell the ship owner, the master, the port installation, the various administrations, Port State Control and the Class society! Strange world, isn't it. Not at all - Total is a rich company - they can pay! The case goes on - in the wrong direction.
On 16 January 2008 the ship owner (GS), the ship manager (AP), the charterer (Total) and the class (RINA) of the M/T 'Erika' were found guilty of negligence and other misdemeanours and sentenced to pay fines and damages by a French court. The master of the tanker and the shore authorities, the latter having refused a safe heaven, were acquitted for various reasons. The flag state, Malta, was not part of the judgement.
Update October 2009
The charterer (Total) has September 2009 appealed against the verdict of 2008. Evidently a charterer or cargo owner cannot be held responsible for alleged, negligent structural maintenance or hidden, structural defects of an oil tanker that later produces an oil spill! The charterer or cargo owner evidently vets many tankers to ensure that it can safely transport the cargo itself but such a vetting is, from a structural point of view, superficial. There is no possibility that a vetting inspector can close up inspect the complete structure in ballast or cargo tanks and determine the actual condition. For that the tanker must be completely gas free, the structure must be accessible and clean (staging or rope access may be required), light and ventilation arranged, etc, etc. In the real world this is only done during the last ballast/complete tank cleaning voyage to and during a scheduled shipyard visit.
Update September 2012
Total lost its appeal and was thus found guilty for not finding the structural defect in the hull prior the incident, the pollution, clean-up and associated damages. The fine is only € 375 000:- but damages/penalties are € 200 million and clean up incl. pumping oil from wreck costs are also € 200 million ... already paid by Total. It was a pity that Total didn't film the wreck so experts could establish why the tanker fractured in the first place.
Lesson to learn
Nowadays no loaded oil tanker ever, officially suffers any structural damages like hull fractures, oil leaks, etc., in service! Reason for this surprising development is that ,if you report such incidents to any port or national authority, the vessel will be turned away as an environmental hazard, i.e. go away and sink somewhere else. So, instead, you report that a crew member has got sick, engine trouble or that some non-vital equipment has broken down and that you thus require a safe haven to fix the problem. Do not ever report a structural hull problem at sea because then you will always be turned away from shore to sink at sea!
And nobody bothers about the fact that there is a safe tanker like the COULOMBI EGG that doesn't spill oil even if it fractures.
Please contact email@example.com of Heiwa Co and make comments! I look forward to your comments how to improve tanker and ferry (and all other ships) safety at sea.