20. DAMAGE TO VESSEL AND DISCUSSION ON DOUBLE HULL VESSELS

Introduction

20.1   From the time the initial grounding took place to the completion of the salvage operation one question which was raised in many quarters was whether the consequences would have been different if the vessel had been of double hull construction. The vessel was thoroughly examined by the Inspectors after she had been taken to the shipyard in Belfast and a comprehensive record taken of the damage which had been sustained. This section of the Report examines the full extent of that damage and discusses whether the pollution would have been substantially reduced, or avoided entirely, if the vessel had been constructed to the standard of a double hull tanker or equivalent. It also examines the regulatory framework for the avoidance of oil pollution under which tankers operate.

Overview of Damage Sustained by the Vessel

20.2   A broad overview of the hull bottom damage is shown in Figure 6 (Image 63k). It can be seen that the most severe damage ran in a continuous band from around the Fore Peak and down the length of the starboard wing tanks. In marked contrast the port wing tanks showed only isolated areas of damage, whilst the majority of the area in way of the centre tanks was largely undamaged or only lightly so. The entire damaged area exhibited pounding damage with the plating dished in between adjacent stiffeners with fractures running along the lines of bulkheads and frames (see Photographs 3 - 6).

Photograph 3 - General View of Bottom Damage to SEA EMPRESS (Image 45k)

Photograph 4 - General View of Bottom Damage to SEA EMPRESS (Image 45k)

Photograph 5 - General View of Bottom Damage to SEA EMPRESS (Image 45k)

Photograph 6 - General View of Bottom Damage to SEA EMPRESS (Image 45k)

A more detailed account of the damage to various areas of the hull bottom follows.

Starboard Bottom

20.3   The most striking aspect of the damage was the extreme degree of deformation to the starboard bottom right along the line of the starboard wing tanks. In this area the turn of bilge had disappeared over virtually the whole of the cargo tank length, and the bottom shell appeared to meet the side shell along a knuckle line some 3 to 4 metres above the keel. Over large areas the longitudinal bulkhead plating had protruded through and then been forced flat against the bottom shell.

20.4   In some areas it was possible to observe earlier fore/aft aligned damage which had been overlain by the plating which had been thrust athwartships. The original raking grounding damage which occurred on Thursday 15 February at the outset of the accident was to a large extent obscured by the overwhelming pounding damage incurred over the ensuing six days to Wednesday 21 February and cannot be entirely separated out. However, the initial bottom raking damage ended at a point about 82% of the length of the vessel aft of the forward perpendicular. This is greater than the 60% figure allowed for in the current MARPOL regulations for a vessel of the size of SEA EMPRESS.

20.5   The damage which caused the initial flooding to No 1 Centre tank, is thought to have been along the line of the longitudinal bulkhead as this was the only damage which was found on the starboard side which would have opened No 1 Centre tank to the sea. The source of the initial flooding to the pump room was found at bulkhead 41. Associated with this damage were fore/aft scores in the hull plating along a line about 6 metres off the centre-line. These two damages suggest that the initial raking damage extended inboard from the starboard side to about 6 metres off the vessel's centre-line. This is illustrated in Figure 7 (Image 37k).

20.6   A section of plating some 2 metres square was missing from underneath No 2 Starboard tank while underneath No 3 Starboard tank there were several areas of plating completely detached from their stiffeners and sagging down. There was also an area some 5 metres long by 2 metres wide which had been folded back along the bottom shell leaving a clear opening in the bottom. The bottom of No 4 Starboard tank had suffered catastrophic damage, the bottom shell structure having been destroyed over almost the entire area of the tank.

Port Bottom

20.7   The turn of bilge over the majority of the parallel mid-body was largely undamaged, apart from the damage to the bilge keel itself. The damage on the port side was concentrated at the ends of the vessel, and usually appeared as discrete indentations. The most notable damage was in way of the pump room.

20.8   The damage to the port side forward of bulkhead 77 was similar to the general pounding damage shown on the starboard side.

Midships Damage

20.9   A massive indentation in the bottom shell structure was present over an area of 10 metres either side of the vessel's fore and aft centre-line. This extended 7 to 10 metres forward from bulkhead 62. The crown of the indentation was some 1.1 metres above the keel while at the crown the bottom plating had fractured along a length of about 17 metres and the faces of the fractured plating were pulled apart by about 150mm and displaced vertically by a further 300mm. The bottom plating in this area is between 18 and 22mm thick. It is thought that this damage occurred as the tide fell on the evening of Monday 19 February when, it is surmised, the vessel must have pivoted about a point somewhere near amidships to achieve the dramatic reduction in her low water draught which the large oil loss at that time indicated.

Pump Room and Slop Tank Damage

20.10   The area of the initial damage to the pump room bottom plating which caused flooding to the pump room was identified just aft and in way of the forward bulkhead 41 on the starboard side. The bottom plating had been pressed upwards with the bulkhead over a length of about 9 metres. In so doing it had split along the base of the bulkhead allowing the ingress of seawater. It is not known to what extent this damage was caused by the initial grounding. However, the damage as inspected is consistent with the limited rate of flooding observed by the salvors and others during Friday and Saturday. The bottom plating on the forward side of bulkhead 41 in way of the slop tank was also pressed upwards, to a greater extent than on the aft side. This split was about 6 metres long, and made the two spaces open to each other and the sea. It is reasonable to suppose that oil escaping from the slop tank was drawn into the pump room via this area of common damage.

20.11   The most notable damage was in way of the pump room on the port side. This damage, which probably occurred on Monday or Tuesday, prevented the pump room from being pumped dry. The damage was less than one frame space away from penetrating the engine room.

20.12   The initial minor damage to the pump room resulted in the atmosphere becoming unsafe when the pump room flooded with seawater and cargo. This meant that lightening operations using the casualty's own cargo pumps could not take place until the pump room had been pumped out and ventilated. This was not achieved until the afternoon of Saturday 17 February. It is arguable that but for this enforced delay to the start of the lightening operation SEA EMPRESS might have been salved on the Saturday without a further loss of cargo. Certainly the lightening tanker was available from 0600 hrs on the Saturday morning.

Damage to Heavy Fuel Oil Tanks

20.13   The first indications of heavy fuel oil pollution came from the remote sensing aircraft whilst the casualty was still in the 'pool'. Prior to refloating it was thought that all the oil had been transferred out of the damaged tank and that the leak had been stemmed. The casualty was not leaking fuel oil on her transit to Herbrandston Jetty. Recognising the potential for pollution from a damaged bunker tank and the amount of bunkers remaining on board, an off-loading vessel was tasked to berth alongside the casualty and remove the majority of her bunkers, thus removing the risk of further pollution. However, during the transfer operation on the morning of 22 February, the casualty's crew began transferring bunkers internally and fuel oil was lost overboard from a tank or piping previously thought to have been undamaged. It was found that both Nos 2 and 4 Fuel Oil tanks on the starboard side were damaged. The damage to No 2 Fuel Oil tank was associated with the damage running along the line of the longitudinal bulkhead (on the port side) aft of frame 43. The damage to No 4 Fuel Oil tank (which is immediately aft of No 2 Fuel Oil tank) was high up on the casualty's side on frame 43. This damage, some 13 metres above the base, was undoubtedly due to contact with the hull made by the tugs during the salvage operation.

Discussion on Double Hull Oil Tankers and Equivalents

20.14   As a consequence of the EXXON VALDEZ accident in 1989, and the unilateral action by the United States Government in favour of double hull oil tankers, there has been a renewed urgency within the marine industry to find a design of vessel which offers the most effective protection against oil pollution in the event of a collision or grounding accident. Several authoritative studies have been carried out into the problem and there is almost unanimous agreement that no one design produces the best results in all the possible grounding or collision scenarios which can be envisaged. There is also general agreement on the following broad conclusions, namely that:

- double hull vessels in low energy (typically low velocity) accidents should not pollute;

- vessels which carry cargo in contact with a single skin (with sea on the other side) will cause some pollution in any accident where a cargo tank is penetrated. However, certain design alternatives will minimize the amount of pollution in some specified scenarios;

- high energy accidents nearly always result in pollution. The relative advantages of various design alternatives in reducing pollution from particular scenarios are highly dependent on the assumptions made in the scenarios.

20.16   One of the original objectives in examining the damage to SEA EMPRESS was to provide data for a subsequent computer analysis to determine how the structure of a double hull tanker (or equivalent) would have withstood similar loads. However, the damage to SEA EMPRESS was found to be both extreme and complex and it was concluded that it was not possible to accurately derive the loading regime which might have caused the observed damage. The plan to carry out a computer analysis of alternative tanker structures was therefore abandoned. Nevertheless a number of scenarios have been selected and utilising the detailed information gained from the inspection of SEA EMPRESS and other areas of the inquiry an assessment, based solely on engineering judgement, has been made to determine how tankers of different structural configuration would have responded to the loadings encountered by SEA EMPRESS.

20.17   To include within this Report the full details of the various scenarios examined is not appropriate, however the findings are given below. The scenarios cover an initial grounding followed by anchoring in the 'pool' (in the same way as SEA EMPRESS) and a further grounding on the shoals off Saint Ann's Head (in the same location as SEA EMPRESS on the Saturday evening). All scenarios envisage tankers of an equivalent size to SEA EMPRESS.

The Double Hull Tanker

20.18   This type of vessel derives its defence against oil spillage, in the event of grounding or collision, by surrounding the entire cargo tank length by a 2 or 3 metre wide void space which separates the cargo tanks from the outer skin of the vessel. In order for an oil spillage to occur the damage has to rupture two skins. In the first scenario with a vessel of this construction, the initial grounding would only rupture the outer skin and there would not be any leakage of cargo. However as the double skin does not encase the pump room that space would be lightly damaged allowing flooding from the sea, however without any oil entering it. The total of this damage would result in an increase in draught which prevents the vessel from proceeding to her berth. It is probable in these circumstances, because the pump room atmosphere is safe, that the flooding is quickly brought under control and the lightening operation (for which a tanker was available from about 0600 hrs on the Saturday morning) is concluded on the Saturday. The casualty is then taken in to the safety of the Herbrandston Jetty during the evening of the Saturday thus avoiding the need to turn the casualty into the weather. The vessel is thus salvaged without any loss of oil. The probability of this outcome would be increased if the pump room were fitted with a double bottom and consequently did not flood initially.

20.19   In the second scenario it is assumed that, for some reason, it is decided to turn the vessel before beginning the lightening operation, that the vessel is turned and anchored in the same location as SEA EMPRESS on the Saturday evening. Then, like SEA EMPRESS, she would certainly be carried onto the shoals off Saint Ann's Head. In these circumstances it is likely that three cargo tanks would be breached with the loss of some 12,000 tonnes of oil. However, in contrast to SEA EMPRESS she could not be salvaged from this location without a prolonged lightening operation, the outcome of which cannot be predicted, but certainly further substantial oil losses would be a real possibility.

The Mid-deck Tanker

20.20   This type of vessel derives its defence against oil spillage, in the event of collision, by protecting the sides along the entire cargo tank length by a 4 to 5 metres wide void space which separates the cargo tanks from the outer side skin of the vessel. The underside of the cargo tank region is unprotected and the cargo is in direct contact with the bottom shell. However, the cargo tanks are split horizontally by an oil-tight deck. The height of the oil-tight horizontal deck is chosen so that in the event of bottom damage the external water pressure should exceed the head of oil in the lower cargo tanks thus forcing the oil to be retained within the vessel. In the first scenario with a vessel of this construction the initial grounding would rupture the void space along the starboard side, at least Nos 1 and 6 Lower Cargo tanks and the pump room. There would be an increase in the draught which prevents the vessel from proceeding to her berth and there would be some oil released. It is possible that the draught of the vessel could be reduced sufficiently, as a result of pressurising the void spaces, to allow the casualty to be taken to the safety of the Herbrandston Jetty. Thus the casualty is salvaged with a minimal loss of oil. Published research data suggests that the total oil loss would have been of the order of 100 to 200 tonnes.

20.21   In the second scenario it is assumed that, for some reason, it is decided to turn the vessel before beginning the pressurising of the void space, that the vessel is turned and anchored in the same location as SEA EMPRESS on the Saturday evening. Then, like SEA EMPRESS, she would certainly be carried onto the shoals off Saint Ann's Head and by the Tuesday would be stranded there. Oil losses up to this time would increase to some 1,000 to 2,000 tonnes. In contrast to SEA EMPRESS, she could not be salvaged from this location without a prolonged lightening operation, the outcome of which cannot be predicted, but certainly further substantial oil losses would be a real possibility.

The Coulombi Egg Tanker design

20.22   At first sight this configuration appears to be a variant of the mid-deck tanker but it differs in three important respects from that type. Firstly the width of the wing tanks is about 50% greater, secondly the wing tanks are divided horizontally into upper and lower tanks with the lower wing tanks dedicated to cargo, and thirdly the upper wing tanks are not only dedicated segregated ballast tanks but also perform the function of "rescue tanks". The Coulombi Egg tanker has an emergency cargo transfer system which allows oil from damaged cargo tanks to be directed into the sound empty upper wing tanks, thus minimising the oil lost to the sea in the event of a collision or grounding. The system utilises the fact that the external pressure from the sea due to the vessel's laden draught will be greater than that due to the head of oil in the damaged cargo tanks, thus the oil in the damaged cargo tanks will be forced into the "rescue tanks". There is only a single skin underneath the pump room. As with the mid-deck tanker no vessel of this type has yet been built but the concept of the "rescue tanks" has been shown to work in model tests.

20.23   In the scenario with a vessel of this construction comparison with the damage found on SEA EMPRESS suggests that the initial contact would rupture the bottoms of the starboard lower wing cargo tanks, Nos 1 andĘ4 Centre Lower Cargo tanks and the pump room. The immediate effect is that the damaged lower tanks would be pressed full as water floods in below the oil and forces it up into the ullage spaces and access trunks. There would be no loss of oil due to hydrostatic effects, but a small loss of oil could be expected due to the forward motion of the ship through the water. The draught of the vessel is increased slightly and there is a small angle of list to starboard due to the small quantity of water which has entered the tanks and due to the flooding of the pump room. In these circumstances it would be possible to proceed directly to Herbrandston Jetty and the salvage operation would be largely circumvented. Published research data suggests that the total oil loss would be of the order of 1,000 tonnes.

Regulations for the Reduction of Oil Pollution from Crude Oil Tankers Following Grounding Accidents

20.24    SEA EMPRESS was required to be designed and built to comply with Annex 1 of MARPOL 73/78 which came into force in October 1983. By these regulations she was required to have segregated ballast tanks arranged to minimise oil outflow and the subsequent pollution resulting from a collision or grounding. The protective area can be placed either in the sides or the bottom of the vessel. These were the standards to which SEA EMPRESS was completed in 1993. Her protective area was concentrated in the sides, against the expectation of collision, in Nos 2 and 4 Port and Starboard Ballast tanks. It is of interest to note that BORGA, which grounded off Milford Haven in October 1995, was designed to the same standard, although in her case the designers elected to obtain the required protective area by fitting a double bottom.

20.25   The fitting of protective areas clearly reduces the risk of damage to the protected cargo tanks. However, one drawback to this mode of pollution avoidance is that in the case where the empty protective segregated ballast tank is ruptured, a significant increase in the draught of the vessel can be expected, accompanied possibly by a large angle of list. This was a critical factor in the SEA EMPRESS accident. It illustrates that in some circumstances the protection of cargo tanks by large void spaces can impede the entry of a casualty to a safe refuge and thus adversely affect the salvage operation.

20.26   The US National Transportation Safety Board in its report on the grounding of EXXON VALDEZ concluded: "....if the EXXON VALDEZ had been fitted with a double bottom, the oil outflow would have been significantly reduced, if not, eliminated." Since double bottoms offer no protection against collisions the double hull was identified as offering the best overall protection against both grounding and collision. In August 1990, as a consequence of that accident, the USA brought into force the Oil Pollution Act of 1990 (OPA90). The most significant requirement of the Act is that new tankers entering the waters of the USA have to be of double hull construction. It also requires that existing tankers, at a date dependent upon their age, have to be retrofitted to double hull standard or removed from service. This effectively means that all single hulled tankers will be excluded from US waters after 1 January 2010 and existing double bottom or double side tankers by 2015. (An exemption from the double hull requirement until 2015 is made in favour of vessels delivering to a lightening operation and/or servicing the Louisiana Offshore Oil Port.)

20.27   In 1992, following the introduction by the USA of OPA90, Annex 1 of MARPOL 73/78 was substantially revised. The amendments came into force in July 1993 and the major amendment was that every new oil tanker of over 5,000 dwt to be of double hull or mid-deck construction or to be of a design approved by the IMO as offering an equivalent level of protection against oil pollution. All existing oil tankers of a size covered by these regulations must comply with the amended provisions no later than 30 years after their date of delivery. This effectively means that by the year 2026 (at the latest) all tankers will comply.

20.28   The MARPOL requirements for double hull tankers, or their equivalent, were enacted in their entirety in the UK Merchant Shipping legislation by Statutory Instrument 1993 No 1680 The Merchant Shipping (Prevention of Oil Pollution) (Amendment) RegulationsĘ1993. Thus, like the USA, the UK is also committed to the requirement that all new oil tankers of over 5,000 dwt should be of double hull construction. However the USA requires that existing tankers shall comply by the year 2015 at the latest whereas the UK and IMO require compliance by 2026.

20.29   OPA90 also requires tanker owners/managers to produce an oil spill emergency response plan, which has to be approved by the United States Coast Guard before the tanker can operate in US waters. One requirement of this, which is recorded in the Code of Federal Regulations is: "Owners and operators of oil tankers and offshore oil barges shall ensure by no later than January 21, 1995, that their vessels have pre-arranged, prompt access to computerised, shore-based damage stability and residual structural strength calculation programs". The managers of SEA EMPRESS had a contract with Lloyd's Register of Shipping, Ship Emergency Response Service (SERS) for this support. This support was therefore available during the SEA EMPRESS accident.

Conclusions

20.30   The SEA EMPRESS accident has highlighted only too clearly some of the fundamental problems which have to be overcome in the design of oil tankers if the risk of oil pollution following grounding damage is to be minimised. However, it has also produced an abundance of information to assist with the development of measures to significantly reduce the risk of oil pollution in similar accidents.

20.31   An analysis of the grounding damage and an exploration of alternative scenarios leads to the following broad findings:

- the easier it is to salvage a casualty the more the risk of further pollution is reduced;

- in coastal waters (where most groundings can be expected to occur) large increases in draught and list following a grounding accident adversely affect the ease with which a casualty can be salved;

- if, because of a large increase in draught, a lightening operation is required in order to salve the casualty it is imperative that the casualty's own cargo pumping system is operable;

- large increases in draught, following grounding accidents, are associated with the rupture of the protective void spaces around the cargo tanks;

- recovering the buoyancy of damaged ballast tanks through pressurisation is more easily accomplished if the ballast tanks are not contaminated with significant quantities of cargo;

- large increases in draught, following grounding accidents, can be avoided by having the tanks adjacent to the bottom skin full of cargo or ballast;

- little oil is lost from breached cargo tanks where the cargo "head" is less than the external positive pressure due to the vessel's draught;

- conversely, massive oil losses can occur if the external pressure due to the casualty's draught falls below that of the cargo "head", as would occur if the casualty were stranded over a low water period;

- of the three generic types of tanker examined only the double hull type has actually been constructed;

- of the three generic types of tanker examined the Coulombi Egg appeared to give the highest probability of avoiding oil losses in excess of 1000 tonnes in a repeat of the SEA EMPRESS accident;

- of the three generic types examined the double hull tanker was the only one offering the chance of zero oil loss in a repeat of the SEA EMPRESS accident;

- the fitting of a double bottom to a pump room would protect the space against flooding and gassing, in the event of a grounding, and have the potential for making a salvage operation easier.