Herald of Free Enterprise Inquiry

The lights should not only indicate in a suitable position on the bridge, but the entire circuit should be designed on a fail-safe basis so that if there should be an electrical failure in any switch circuit the system would indicate danger. Indicator lights should be fitted to all superstructure doors, such as passenger access, bunkering and storing doors. It is desirable that a dedicated door alarm panel be fitted, similar to that commonly used for watertight doors. Finally, the condition of superstructure door indicator lights should be logged before departure just as is the condition of the watertight doors below the bulkhead deck.

We have no doubt that closed circuit television monitoring of all superstructure doors, is well worthwhile and should be fitted.

It follows that if ships cannot shut their doors at a particular berth, because of the design of that berth, then alterations should be made to the berth. Alterations have been made to some berths, but it is considered by the Court that all berths on U.K. routes should be altered so that ships with clam shell doors or normal stern doors are able to shut their doors before leaving the berth. It is more difficult for ships with visor bows, which must pull back somewhat from the ramp before they can close the visor. But the fact that the visor bow is open is obvious to the Master. Indeed the ship cannot be navigated until the bow visor has been lowered. This is, therefore, not a serious problem provided the inner door is shut before leaving the berth. It is desirable that each berth should then carry an approval certificate specifically listing the ships which can operate from it, and can shut their bow and stern doors without moving from the berth. Parliament may wish to consider whether it should be an offence for a ferry to leave a berth before its bow and stern doors are closed.

The Court concludes that mechanical, pneumatic, electrical or hydrostatic draught gauges or indicators should be a requirement for Ro/Ro passenger ferries, using types specifically investigated and approved by the Department. These gauges or indicators should be fitted to give readout at the aft and forward loading positions and on the bridge. Desirably they should indicate the aft and forward draughts and the midships draught. The latter should not be a mean of the fore and aft draughts but should be a mean of sensors, port and starboard, giving a true draught at the loadline mark. Draught gauges or indicators should, if possible, be suitable for interfacing with a loadicator so that if weight information for vehicles is fed into the loadicator a running automatic up-date of the ship’s condition could be produced. It is recommended that the loadicator be in a suitable central position with, if possible, work stations at the two loading stations and on the bridge.

The practice of using one metric tonne for the all-up weight of the average car with luggage, fuel, and personal effects, is outmoded. A nominal weight of 1¼ metric tonnes should be adopted. This figure should be reviewed regularly by the Department on a large random sample basis, say, every five years. Similar considerations apply to coaches, where the current Townsend Thoresen practice of allowing 14 tons per coach should be adopted. It is desirable that the Department should conduct an analysis of coaches, particularly of the long distance double decker type, in order to determine a more accurate basis for the nominal weight. This nominal weight should then be used in making up the cargo predictions. Finally, there is the very important question of freight vehicle weights. Partly because the permissible weight limits on the Continent are higher than in the United Kingdom there is an advantage to be gained in under-declaring the weight of freight vehicles proceeding from the Continent to the United Kingdom. It is not enough to say that this will only apply to vehicles coming into the United Kingdom because if a vehicle is leaving the United Kingdom for a long Continental run, there is also an advantage for the British operator to load to the Continental limit. Checks at Dover on over 100 vehicles coming to the United Kingdom showed that on average the actual weight exceeded the declared weight by approximately 13%. It is considered that these checks should be extended, particularly covering the present gap in knowledge of the weight of outgoing vehicles. The checks should not be on the basis of weighing obviously likely vehicles, but on a purely random basis. To avoid criticism of the statistical validity of such checks, a considerable number of vehicles should be covered, say 500 either way. Such numbers, in conjunction with a random choice, should produce a reliable margin for addition to the waybill declaration. Meanwhile we consider that an addition of 13% should be made to declared weights when making up anticipated cargoes. This does not, however, replace the necessity to read draughts accurately to prevent overloading, nor does it replace the need for weigh bridges and individual weight measurements for every heavy vehicle loading onto a Ro-Ro ferry. Its main value, in the context of accurately read draughts, is to allow reliable determination of the ratio of weight being loaded on the upper and lower decks of a two-tier ferry. This ratio can be crucial, as was shown by evidence of Mr. Taggart. Finally, once a comprehensive weigh bridge system is instituted the Department should review the statistics of declared and actual weights and produce an update of such margins every five years.

In view of the uncertainty as to the actual weights of many freight vehicles every effort should be made to persuade, or even require, Port Authorities to provide rolling weigh-bridges, possibly of the loadcell type, where all freight vehicles coming into a port will be automatically weighed and issued with a weight certificate, to be attached to the vehicle. If such weigh bridges were fitted in way of each ramp it would be possible to pass this information electronically to the loadicator, thus updating the calculated condition of the ship if the loading officer punched in vehicle locations. The Court concludes that the Department should encourage Operators to investigate immediately the technical and practical aspects of the proposal. As the calculation of fluid KG (or GM) of a Ro-Ro vessel prior to departure should be obligatory such a system would be of real assistance in carrying this out.

As a matter of urgency, self-contained, maintained emergency lighting units of a type approved by the Department should be fitted to all Ro/Ro passenger ferries in suitable numbers and in such places as may be advised by the Department. It is essential in our view that such units should be watertight - in other words, if the unit should be submerged it should remain alight until the batteries run down. Furthermore, a regular routine check on the condition of this emergency lighting should be a requirement. This could perhaps be a weekly inspection by the ship’s electrical officer followed by logging that all was in order.

If laminated safety glass is to be used in windows in way of lifeboat and embarkation stations, it is clear that they should, if possible, be fitted in push-out or centre line hinged window frames in situations where rapid exit through the windows may be necessary. In other areas it is essential to stay with heat treated toughened glass, with the minimum grain size compatible with the other desired characteristics. In this respect there is clearly a need for a good standardised industry approach to escape windows. These should be reliable and uncomplicated and openable from either side.

The Court thinks that in general, the design of slab sided vessels should be discouraged. The main objections to them are:- 1. They eliminate outboard fore and aft escape routes. 2. Embarkation stations often have to be on the top deck, which is very high above the water. It is in the context of current design that the importance of being able to break side windows arises, not only if a ship is on her beam ends, but also to allow the possibility of escape in extreme circumstances when the ship is upright. Athwartships doors should be provided at recognised intervals; and thought should be given to providing access to them with the ship at significant angles of heel, say in excess of 20°. Additionally, it is a matter of some concern that passengers should be able to recognise readily escape routes and doors allowing exit athwartships from these long passenger compartments. Prominent labelling of decks, exits, lifejacket stows and muster stations is important. The Court considers that designers and the Department should give attention to these matters. Another matter of concern is the efficiency of emergency exits from accommodation below bulkhead deck. A steward, Mr. Butler, heard water running down the stairs from G deck to the lower accommodation. This must have been going over the sill from the vehicle deck which was 200 mm in height. Consideration should be given to whether such sills should be higher, say 600 mm. The HERALD had two main passenger compartments below the bulkhead deck, between frames 105-118 and between 118-132. Each of these was fitted with two means of escape in accordance with the P.S.C. Regulations 1980. However, there was a watertight door between the two passenger compartments and this was often kept open for convenience and access. If this door was open during the casualty that was not causative. Nevertheless such a door should be kept shut at sea and not used as an escape route. This is a matter which should be implemented immediately. Furthermore, there should be immediately a general safety audit of means of escape from passenger compartments below bulkhead decks in existing ships to ensure that:- (a) they comply with the relevant Passenger Ship Regulations, and (b) that the Regulations themselves are being applied properly and rigorously. A single means of escape from spaces below the bulkhead deck is dangerous. Supplementary escapes should take into account the possibility that collision damage or submergence may render them useless.

While suggestions have been made as to the desirability of finding some means of bridging these shafts, it is the view of the Department of Transport that this is not really practicable. the Court is not convinced by this. It considers that simple methods of bridging should be developed. For example, split deckhead panels hinged in way of the sides of a longitudinal passage could be arranged so that with the ship on her beam ends one of these panels could be brought down and made to bridge the gap. It would be self-supporting if it extended fore and aft beyond the sides of the transverse alley way. There may well be better methods, but some such simple device should be developed and made a requirement.

There should be a requirement in the PSC & S Regulations for standard envelope curves, at least for level keel and trims by the stern and by the head of, say, 0.4% and 0.8% of the length of the ship. In view of the importance of this point, the Court considers that requirements should be imposed upon all existing ferries for the provision of such initial trim curves. These should be implemented within a short time to prevent operation at unsuitable (and possibly dangerous) trims and vertical centres of gravity. The Court also considers that all existing Stability Books for Ro-Ro passenger ferries should be reviewed in detail by the department. Furthermore, during the review of existing books they should be examined to see whether they provide adequate information for the assessment of safe loading in respect of factors such as the tabulation of vertical centres of gravity for cargo on all vehicle decks. A standardised vertical centre of gravity should be used for each type of vehicle.

Not only should the SPIRIT class vessels be re-inclined, but all existing ferries that have not been re-inclined within the last 4 years should be re-inclined as a matter of urgency to assess their current lightship weights and centres of gravity. Thereafter they should be re-inclined at every special survey. These figures would then be available for use in the production of reliable trimmed envelope curves.

The Department should state that trims should never exceed certain limits, or alternatively and preferably, hydrostatic data for large trims should be included in the Stability Booklet, but noted as being for harbour guidance only. Furthermore, the Stability Book should emphasise the need to keep a running check upon metacentric height at the relevant trims during the unloading/loading operations. The Stability Book or Approved Program must apply to all operating conditions, not only to seagoing conditions.

The Court recommends that detailed investigations and model tests should be carried out with a view to increasing the required distance from the margin line to the bulkhead deck in new design Ro-Ro passenger ferries to perhaps 1 m. The model tests should be ballasted to give the waterline at the margin over a range of 'damage' positions fore and aft and with the relevant openings in the superstructure representing statutory damage. Tests should be run in irregular sea conditions for significant wave heights. The present definition of the margin line (76 mm below the bulkhead deck) derives from conventional passenger ships and has no inherent relevance to ships with no subdivision above the bulkhead deck. It should not apply to them.

Immediate consideration should be given to phasing out vessels built under the 1965 rules unless they meet or can be modified to meet, at least, the 1980 standards in these respects, as they may be substantially less safe than modern ships. If they do not meet them and cannot be so modified, a finite and short term should be put on their lives.

Alternatively there are lines of investigation that should be pursued urgently with the object of finding methods of improving the survivability of Ro-Ro passenger ferries. 50.1 The vehicle deck may be subdivided transversely by permanent bulkheads with doors. This method would bring a Ro-Ro ferry into conformity with Rule 24 of the 1984 Regulations. The ship would then parallel normal passenger ship construction. However, such transverse bulkheads, if permanent, even with doors, could constitute a grave penalty against the commercial operation of a ferry for its primary purpose. It is concluded that a feasibility exercise should be carried out to investigate the practical details of such divisions and their effect upon commercial operations. In the case of train ferries this method would be prohibitive. Trains are loaded in rakes and then compressed against buffers. No transverse division whether permanent with doors or portable could be accepted. The Court heard evidence from Mr. Taggart illustrating the potential effect of transverse bulkheads on the HERALD. Three bulkheads above the bulkhead deck were envisaged. Damage was assumed alternatively in way of two of these bulkheads. Considerable improvements resulted. Undoubtedly transverse bulkheads in the vehicle deck would effect a considerable improvement in the safety of these vessels, but the thrust of any investigation must be directed towards operational practicability. 50.2 An alternative possibility for subdividing the vehicle deck would be to fit portable transverse bulkheads. In one form these could be top or side hinged doors. Such bulkheads are fitted to deep sea Ro-Ro freight vessels. However with deep sea vessels the need for a minimal turnround time is not as pressing as with short sea ferries, especially on routes such as Dover/Calais. Loading has to proceed sequentially from the far end of the ship towards the ramp and this may introduce trim problems. Portable transverse bulkheads would increase the turn-round time and result in a small loss of stowage space. Alternative types of portable bulkheads could be envisaged, perhaps vertically sliding, housing below the bulkhead deck; but these would only be possible in the event that the bulkhead deck was not a strength deck. Other possibilities, such as vertical roller shutter bulkheads or alternatively sideways rolling shutter bulkheads should be investigated. A detailed study should be made of all possible types of portable bulkheads, both structurally and in terms of their practical effects including concomitant vehicle stowing disadvantages as mentioned above. If fitted, either these or transverse bulkheads with doors could afford a major increase in the survivability of Ro-Ro passenger ferries. It is conceivable that arrangements could be developed to keep loss of space and vehicle handling time within acceptable limits. 50.3 Another possibility is to divide the vehicle deck with side longitudinal bulkheads, the space outboard of these being themselves subdivided transversely. The Court heard evidence on this proposal. This was in the context of the HERALD herself. The inboard side of the line of casings was on the B/5 line in the midships area. Bulkheads were envisaged extending fore and aft until reaching the forward and aft compartments, with three transverse bulkheads in the main length of each outboard space. Suitable access doors were postulated. On the assumption that two compartment damage did not penetrate the B/5 bulkhead, the result was a range of stability and GZ giving full compliance with all the required damage criteria. This proposal is practicable and satisfactory in the context of a vessel with twin casings. It would not be applicable to vessels with centreline casings. The arrangement would be likely to be dangerous should damage penetrate beyond the B/5 line. Mr. Rogan introduced evidence from 1974 I.M.O. Regulations on subdivision and stability concerning the statistics of damage penetration from 296 penetrations. This indicated that approximately 45% of penetrations exceeded the B/5 line. Mr. Rogan gave his opinion that this could produce a more dangerous situation than would have existed without the longitudinal bulkheads. The Court agrees with this view. However the proposal has some attractions and there is an urgent need for investigations into the effect of longitudinal bulkheads on a range of vessels. These investigations should include depth of penetration probability statistics and a definition of the risks involved with different configurations and penetrations.

In conjunction with all such arrangements, detailed consideration should be given to access openings from the vehicle deck. Casings and trunks usually contain staircases leading down to spaces below the bulkhead deck and up either to another vehicle deck or to accommodation. In the HERALD casualty the first indication of trouble was water coming down the stairs into the forward drivers' accommodation. There was little to stop water going into the lower levels of accommodation. This point was underlined by Mr. Taggart in his evidence about attempts to obtain agreement on Regulation 8 of the 1974 Amendments to SOLAS Convention. Questioned regarding the residual range of stability and the contrast between the maximum heel figure of 7° proposed by the United Kingdom and others as against the 20° proposed by the U.S.A. and the U.S.S.R., Mr. Taggart said that to achieve the latter angle would be extremely difficult. "I would think to achieve 20° in most vessels there are so many problems in passenger vessels with accesses within the ship to below decks that effectively your curve is terminated because of progressive flooding problems well before 20° of heel. That is not 20° angle of heel from upright; that is 20° from the equilibrium position . . . So that you are talking about possibly 27° or more. You would require to keep all your openings safe so that no progressive flooding could occur within say 27°. It would be a major problem . . . It is most important in these things not to impose criteria which are impossible to achieve." It should not be very difficult in Ro-Ro passenger ferries to ensure major improvements in these respects. Sill heights should be increased appreciably. Wherever possible access to spaces such as the engine room and below bulkhead deck passenger accommodation should be without intermediate access to the bulkhead deck. Exits in the casings from this vehicle deck should lead only upwards. Access to below bulkhead deck spaces, except where through watertight doors kept shut at sea, should be from the deck above the bulkhead deck. The Court does not agree that it would be a major problem to achieve substantially higher downflooding angles. It would seem to be relatively easy and achievable even on many existing vessels. Immediate attention should be given to this point.

This conflict between I.M.O. and other Regulations should be investigated and clarified as a matter of urgency.

There is a fourth area which is apparently missing namely:- Improvement in the prevention of water reaching the vehicle deck in the event of damage occurring in realistic seagoing conditions. The latter area is as important as the other three and logically should precede them. This point carries over to Task 4, namely survival capability. The tank test programme should cover a wide spectrum of possibilities such as the systematic testing of survivability in waves, with varying distances from the margin line to the bulkhead deck and the provision of external add-on buoyancy. In particular the validity of the present survival criteria should be checked experimentally, in a range of irregular significant sea heights with the enclosed superstructure opened by side damage. It should be emphasised that this Formal Investigation is into the disaster to the HERALD and into possible means of improving the safety and survivability of Ro-Ro passenger ferries. In this context Task 1 should concentrate on Ro-Ro passenger ferries. It is essential that all calculations, including those for residual stability and freeboard, should be in the context of a range of initial trims. The casualty analysis and relative safety proposals appear to have been covered largely by the Norwegian submission to I.M.O. submitted in evidence during the HERALD Investigation (November 1983 Paper SLF/20). Between the evidence based upon Lloyd’s Register’s work and the Norwegian paper, this section would appear to be largely unnecessary. It is difficult to see how the relative safety of new Ro-Ro ships can be compared statistically with existing ships. Risk assessment is also likely to be upon an inadequate statistical basis. This study does not appear to cover actual tank testing but finishes by preparing specifications for a programme of tests. Thus it seems unlikely that there will be any tank test information available for the September 1987 meeting of the SLF Sub-committee of I.M.O. It is considered that the tests should impose statutory damage through portable side plates in the superstructure located at, say, midships and near the quarter points. The model should then be ballasted so that (a) the waterline is at the margin in way of the ‘damage’ location for any given distance between margin line and bulkhead deck. (b) that the metacentric height is adjusted to be alternatively 0.05 m or 0.10 m; and (c) finally, that the vessel be tested statically on a number of headings in irregular sea conditions using a suitable spectrum with a range of significant waveheights. Finally, no target date for the completion of a model test programme is given. It is of considerable importance that this be as soon as physically possible in view of the need to consider the effects of the results upon ferry design. It is essential that the research programme convince Government and the Public that its conclusions are objective and at arm’s length from purely commercial considerations. In particular, the merit rating mentioned in 2.4.4. and 2.5.3 of the programme, requires clarification as to its nature. Is it, for example, to be an operator’s assessment of least possible inconvenience or is it to apply strictly to the safety of the vessel or will it be based upon an objective combination of these and other aspects?

This Court need say no more than stress the need for: (a) Clear and concise orders. (b) Strict discipline. (c) Attention at all times to all matters affecting the safety of the ship and those on board. There must be no “cutting of corners”. (d) The maintenance of proper channels of communication between ship and shore for the receipt and dissemination of information. (e) A clear and firm management and command structure.

It is the view of this Court that the time has now come when the whole philosophy of carrying lifeboats on ships which are never far from land should be reviewed by the International Maritime Organisation. This is a topic which requires much careful consideration before any International agreement is likely to be reached. The Court would, however, draw attention to three particular aspects of this question, two of which relate specifically to Ro-Ro ferries. (1) All ferries have heavily constructed belting which protrudes from the ship’s side. Although they have farings in way of the boat lowering positions the guard rails can catch the gunwales of a lifeboat, particularly if there is any sea running, and cause it to roll violently. (2) The means of release gives rise to difficulty and sometimes to moments of danger, as for example if both ends are not released simultaneously, with the result that the boat is “hung up” at one end or the other. The heavy blocks are a danger to occupants of a boat as they swing to and fro on the ends of their falls. (3) There was some evidence that the height of the embarkation deck should be as low as possible. There are practical reasons for this, but in an emergency it seems to be against human instincts to go to a lower deck than the one a person is on. Passengers are more likely to want to move to a higher deck. There are obvious advantages in drastically reducing the number of lifeboats carried and thereby saving space and reducing weight. In their place a large number of liferafts could be carried with advantage. The Court recommends that the Department should give a lead in initiating discussions.

There should be a requirement that the departure draughts must be entered in the deck log book as well as the Official Log book. The only practical way of enforcing such a rule would be for the Department to initiate random spot checks and for there to be heavy penalties for an infringement of the Regulations. The Court recommends that further consideration be given to this topic.

Consideration should be given to enlarging that regulation to include every occurrence which is potentially hazardous to the ship or to any person on board.

The court draws attention to this suggestion and expresses the hope that serious consideration will be given to it.

In the course of considering all the related problems, further consideration should be given to the question of appointing counsel to the tribunal, and not on behalf of the Secretary of State, so that he can be seen to be wholly independent of the Department.