2011X59 Loss of the DERBYSHIRE - Official Report

Some 69 individual items of wreckage were visited for close-up viewing which generated some 200 hours of prime video imagery. 12 individual items were visited to obtain macro images of fracture surfaces yielding 302 individual specific shots. Subsequent analysis of the imagery shows that the majority of the vessel has been specifically identified, viz. 
~100% of the centre line of the vessel, ~ 85% of the port side deck plating, ~ 80% of the starboard side deck plating, the bow, stern and accommodation, ~ 23 items of hatch cover wreckage were found of which 13 complete were covers and 10 were pieces. Reconstructing the 10 pieces by area resulted in the identification of a further 4.2 of the remaining 5 covers. The wreckage field must contain in excess of 2,000 items.

Conclusions - Taking all the survey evidence into account we conclude that:

• The most likely cause of the loss of the DERBYSHIRE was that the spaces forward of the collision bulkhead at Frame 339 became more or less flooded over a period of time initiating the sequence of events.
   
• This flooding resulted in a substantial reduction in the freeboard of the vessel at the forward end and a reduction in the ability of the vessel to 'rise to' the waves in the Typhoon conditions.
   
• As a consequence, the forward cargo hatch covers were subjected to considerably increased wave heights and dynamic pressures. These heights, in excess of the design parameters, resulted in the failure of the hatch covers and subsequent foundering of the vessel.

The Principle Findings

• The vessel was hove to in the most dangerous sector of a major Typhoon system where high waves prevailed and very high waves of low probability of occurrence could have been encountered exceeding what might be considered the 'normal' design envelope of the vessel.
   
• The vessel was almost fully laden with high density cargo with 2 of the 9 holds empty.
   
• The destruction of the vessel as seen on the seabed was caused by the IMPLOSION/EXPLOSION MECHANISM. It accounts for the near total fragmentation and disruption of the vessel structures. No other mechanism could wreak such havoc with a ship's structure except explosives, there was however no evidence of pure explosion damage, burns or scorch marks.
   
• Violent implosions / explosions cannot occur unless compartments remain with strong and intact boundaries down to the requisite collapse depth. In this case, by calculation, such depths would be between 25m and 100m for primary structures and, as seen, the failures followed the predictions.
   
• Every piece of the cargo section, between Fr65 and Fr339, has been violently and destructively affected by the Implosion / Explosion Mechanism either as a primary failure or secondary effect.
   
• The cofferdam structures at bulkhead 65 were found integral, i.e. the bulkheads (Frs64/65) were still joined by the horizontal and vertical diaphragm structures between them, which means that no hull separation in the vertical plane occurred between the two bulkheads.
   
• At the stern, the slop tanks immediately aft of the Fr64/65 cofferdam, tanks in the engine room double bottom, and most probably air receivers and other tanks within the engine room, also imploded causing major destruction.
   
• The bow section was clearly largely unaffected by implosion / explosion.

• It was therefore largely pressure compensated, i.e. full of water, as it left the surface. This water entered the bow through the various open ventilators and the unsecured hatch found in the survey.
   
• The filling of the bow would have been a relatively slow process over a period of hours and this, in conjunction with the hatch covers having failed in the holds, directly contradicts those loss scenarios pre-supposing a rapid initiation and sinking by capsize.
   
• Model tests previously carried out showed that the normal forward freeboard of the vessel provided barely sufficient reserve buoyancy in a heavy seaway to protect deck fittings and hatches from green sea impacts. In a Typhoon, with forward compartments flooded, the vessel could not rise to the waves and waves could sweep over the bow.
   
• All the hatch covers were found within the wreckage field and indications are that all had been firstly driven into the holds by external pressure, one at least by dynamic impact.
   
• The macro imagery of the many fracture surfaces examined showed that brittle type failures abounded throughout the wreckage occurring in both 'A' Grade and 'E' Grade materials. The diversity and evident violence of these failures could only have been as a result of the implosion / explosion mechanism.
   
• The above findings, coupled with the disposition of the wreckage, showing its relatively small spread on the seabed, provide confirmation that the vessel was structurally complete and without failure of the main hull girder at the sea surface.
   
• The more or less even distribution of the cargo throughout the wreckage field and the inversion of double bottom structures makes unsafe any conclusions which may be drawn regarding the possible capsize at the surface.
   
• The Anomalies, being the unusual set of sonar targets some 5 kilometres form the wreckage field, were found to be natural features of volcanic origin and therefore not pertinent to the loss.

The Loss Scenario Analysis concluded that the following were not the cause of the accident:

Deck cracking at Frame 65, deck cracking at mid-sections, torsional weakness, hatch cover attachments, fore deck corrosion and fracture, cargo shift / liquefaction, propulsion loss, rudder loss / steering failure, explosion and / or fire in the engine room, "pooping" from forward waves, "pooping" from running with the sea. No "unforseen" loss scenario was found.

Hatch cover failure was the final consequence which caused the loss. The slow filling of the bow prior to their failure indicates that this was not however the initiating event. The 14th scenario investigated - Flooding of forward spaces - was considered to be the initiating event of the loss and this scenario describes the cause of the loss. The subsequent hatch cover failure was the final consequence of the loss of the vessel.

We will never know exactly how the vessel sank since there were no surviving witnesses. The best we can do is to postulate a sequence of events and translate this into a more understandable form which draws the various threads of fact and augment them together in a concerted way. We therefore offer the following:

As the Typhoon approaches and sea states increase, the Master turns head to the seas in the prescribed manner, keeping the seas just off the bow. The vessel begins to take green seas over the bow of increasing frequency and intensity. The unsecured fore deck hatch cover is carried away and the Bosun's Store space fills with every wave over the bow, probably becoming full in less than one hour. Because the hatch was hidden from direct view of the bridge by the mast and other fittings and, not least, the continual green seas breaking across the deck, the Master cannot see this happening. Because of the increasing pitch motions and wind and spray over the deck, the Master also cannot see that the bow has lost freeboard. (Even if he could see this problem, the waves breaking over the deck would make it unsafe for the crew to attempt to intervene). In the darkness, the crew could not switch on the fore mast floodlights since the switches were in the fore peak stores.

As the main force of the Typhoon approaches and the seas increase even more with substantial green seas now crashing on the fore deck, the Master heaves-to matching, as far as possible, the wave drift forces with his main engine and keeping the waves just off the bow on the starboard side. Damages now begin to occur to deck fittings. Ventilator mushroom heads and floating ball types are carried away or rendered open to the sea. The fore peak ballast tank for example begins to take in water with every wave and, over 12 hours is more or less filled. The vessel has now gone down by the head by some 2.5m but this cannot be realised by the Master due to the ferocity of the wind and seas which are now crashing and whipping across the decks.

The vessel would now be more sluggish to rise to the waves and 25m waves would be sweeping over the bow and dissipating their energies along the deck over the first three to four hatches. It is most probable that at this time the vessel ìtook a bad oneî, a high, steep wave, perhaps even a very high wave or group of such waves of low probability of occurrence, combined with the vessel still driving its bow downwards on the back slope of the previous wave. The wave crashes over the starboard side of the bow and destroys the starboard hatch cover of No. 1 hold, fracturing it in the middle and driving it into the hold followed by hundreds of tonnes of water. The port side hatch cover follows, either at the same time or with the next wave. In less than 1 minute, with wave encounters now at about 13 second intervals, No. 1 hold fills with 10,000 tonnes of sea water. All forward freeboard and reserve buoyancy is now lost and waves roll un-impeded and without breaking over the forward end and along the deck, as if up a shallow beach, some even approaching the bridge front.

The foremast and forward doghouse are carried away and water cascades down into and along the duct keel, filling the cofferdam. If not already breached, the hydraulic machinery space bulkhead fails, imploding the space and damaging the deck over. The ship is now foundering. No.2 hold hatch covers collapse under the wave pressure and the hold fills, once more in less than a minute. As she slips below the seas, the remaining hatches are driven inwards by wave pressures which become increasingly violent as the 'beaching' angle increases with the bow going down. Before the stern goes below the seas, implosions/explosions are already taking place at the bow. The fuel tank, already containing fuel oil and sea water, implodes internally and relieves the increasing pressure on its bulkhead 339 boundary, the air in the ullage space exploding this bulkhead outwards at the top. No. 1 hold tanks begin to implode and the whole process dominoes along the length of the vessel in rapid fashion with shock waves traversing the structures weakening them and making them more susceptible to collapse towards the aft end.

The stern is the last section to leave the surface, probably now at an angle of 45 degrees down towards the bow and rolling to port. Implosions at the forward end of the slop tanks rupture the engine room structures, the pump room and No.9 hold section tear out the double bottom, aided and abetted by the weight of the engines and auxiliaries. The stern immediately fills, sinking below the surface. The accommodation, already seriously damaged, separates below the surface. Only a few minutes have passed since No.1 hatch covers failed and there was not enough time to send a distress signal. The ship is now breaking broadly into the sections seen on the sea bed and hydrodynamic forces take over. It probably took about 1 hour for the ship to reach the sea bed.

Recommendations

Taking into account not only the negative aspects of our findings but also the positive, we include a discussion on a number of possible changes in vessel design, rule and regulation which we believe will serve to improve not only the safety of bulk carriers but ships in general. These include:

Formal safety assessments, design for extreme weather conditions, changing design philosophies, double skin plating in way of holds, loss of steering ability, fore end flooding, fore deck and fore end spaces access and hatch covers.

More detailed proposals are made regarding improvements in both fittings and safety systems including ventilator and air pipe design, positioning and protection, spurling pipes, tank sounding systems, bilge alarms, means of emergency pumping, access hatches and fore deck lighting. Obviously these proposals will require detailed discussion and adoption into the structure of the world maritime community but it is hoped that they will be seriously considered on the basis of evidence and the name of DERBYSHIRE, and her crew, will live on in the literature as the promoter of improved ship safety. It is also hoped that the methodology and technology utilised on the wreckage of the DERBYSHIRE will show the worldwide maritime community that no ship can be lost, however badly damaged it may be, without the prospect of location and survey as to the cause of the loss.