(1) KASTOR NAVIGATION CO LTD (2) ATLANTIC BANK OF NEW YORK v (1) AGF M.A.T (2) AXA GLOBAL RISKS (UK) (3) SOCIETA ITALIANA ASSICURAZIONIE REASSICURAZIONI S.p.A (4) S.A.S.A. S.p.A (5) AXA CORPORATE SOLUTIONS (Formerly known as AXA GLOBAL RISKS) (6) UNITAS GJENSIDIG ASSURANSEFORENING [2002] EWHC 2601 (Comm)

IN THE HIGH COURT OF JUSTICE

QUEENS BENCH DIVISION

COMMERCIAL COURT

[2002] EWHC 2601 (Comm)

Case No Case No: 200 Folio 1196

Venue Royal Courts of Justice, Strand, London, WC2A 2LL

Between

(1) KASTOR NAVIGATION CO LTD (2) ATLANTIC BANK OF NEW YORKClaimants

(1) AGF M.A.T (2) AXA GLOBAL RISKS (UK) (3) SOCIETA ITALIANA ASSICURAZIONIE REASSICURAZIONI S.p.A (4) S.A.S.A. S.p.A (5) AXA CORPORATE SOLUTIONS (Formerly known as AXA GLOBAL RISKS) (6) UNITAS GJENSIDIG ASSURANSEFORENINGDefendants

Before

Mr Justice Tomlinson

Mr Stephen Hofmeyr QC and Mr Andrew Baker (instructed by Ince and Co.) for the Claimants

Mr Steven Berry QC (instructed by Holman Fenwick and Willan) for the Defendants

Hearing dates : 1,2,3,4,8,9,11,15,16,17,18,22,24,25,29,30,and 31 July 2002.

Date 4 December 2002

[1]Mr Justice Tomlinson : This is a claim against underwriters in respect of the total loss of a ship. "Kastor Too" was a steel-built four hold geared motor bulker of 17,665 tonnes deadweight, 148 metres in length built in Japan in 1977. The First Claimant was the registered owner of the vessel. The Defendants subscribed to a contract of marine insurance contained in a slip policy on the MAR 1991 Form insuring the First Claimant's interest in the hull and machinery of the vessel against marine risks for a period of 24 months at 30 December 1999. The vessel was valued at US$3M. The Defendants insured 80% of the risk. The insured perils included fire, explosion and perils of the sea. The Second Claimant was mortgagee of the vessel and the assignee of and loss payee under the policy.

[2]On 29 February 2000 the vessel sailed from Aqaba bound for Vizagapatnam laden with a cargo of rock phosphate in all four holds. At about 1420 hours on 9 March when the vessel was in a position between the island of Socotra and the coast of South Yemen a fire was discovered at the purifier flat level in the engine room. The fire was observed by the Chief Engineer and the Third Engineer through a window in the engine room control room where they then were. The Third Engineer took a 5 kg CO2 fire extinguisher from the control room but he could get no closer to the fire than the stairway on the starboard side of the engine room because of the smoke and flames so he emptied the extinguisher from there in the general direction of the fire to no effect. There were no further efforts to fight the fire and the engine room was abandoned within two or three minutes of the initial observation of the fire. Although the emergency fire pump located in the steering gear room was started the main fire pump at bottom plate level was not. The emergency fire pump was used for boundary cooling. After closing of all openings in the engine room the fixed CO2 system was activated. This appeared initially to have the effect of reducing the amount of smoke issuing from the funnel area. This effect was shortlived. Within thirty minutes smoke escaping from the funnel area and the starboard engine room ventilator began to increase significantly. The crew abandoned ship and took to the two lifeboats. All personnel including the Master and Chief Engineer, the last to leave, were in the boats by about 1700. Between 0500 and 0600 hours the next morning, ie, about 15 hours after the initial observation of the fire, the vessel sank stern first.

[3]The ship would not have sunk with the engine room alone flooded. In order for her to sink at least two compartments must have been flooded, the only realistic contender here in addition to the engine room being No.4 hold, which is immediately forward of the engine room bulkhead. A steel ship constructed in accordance with the Load Line and other conventions in force when this ship was built should not sink as a consequence of a fire in the engine room. For a steel ship constructed in this way to have sunk as a consequence of fire in the engine room within only 15 hours or so of the outbreak of the fire would be very remarkable. That is no doubt why the underwriters have declined to pay the Claimants' claim for the total loss. Although the underwriters adduce no positive case as to the cause of the sinking, no one could have been in any doubt as to the underwriters' suspicion that the vessel had been scuttled.

[4]This action was begun on 1 November 2000. Initially the Claimants claimed for an actual total loss, allegedly caused by fire and/or explosion and/or perils of the sea. I leave out of account a subsequently abandoned allegation of crew negligence. I can also henceforth omit reference to perils of the sea. The Claimants' only case was that the fire and/or explosions caused a fortuitous entry of seawater, so reliance upon perils of the sea adds nothing. Initially, underwriters neither admitted nor denied that there had been a fire but denied that, if there had been, it had caused the sinking of the vessel. By an amendment to their pleadings introduced in August 2001 the Claimants claimed additionally or in the alternative for a constructive total loss of the vessel. They alleged that prior to her sinking the vessel was so damaged by fire that the cost of repairing the damage would have exceeded the insured value. Initially, the underwriters neither admitted nor denied that the vessel had become a CTL. Before the trial the underwriters had put in a report from their expert marine engineer, Mr Todd, which accepted that the likely cost of repairing damage of the type which is likely to have been sustained by "Kastor Too" prior to her sinking would have appreciably exceeded US$3M. However, underwriters continued to deny that the vessel had become a CTL, since their admission as to the cost of repairs related only to the scintilla temporis prior to sinking, at which time they said the vessel was doomed to become an actual total loss. Further points emerged at trial as to the ability of the Claimants to claim for a CTL having initially claimed for an actual total loss.

[5]The battle lines which were eventually drawn, either by the time the trial began or at any rate during it, were as follows. The Claimants' primary claim was for a CTL caused by fire and/or explosions. Obviously that ranked logically before their alternative claim which was for an actual total loss also caused by fire and/or explosion. The underwriters admitted that there had been a fire in the engine room starting at about 1420 hours. Their fire expert, Dr. Bound, was instructed to prepare his report on the assumption that the fire started accidentally. In his final address Mr Berry QC for the underwriters observed that, having accepted that the fire occurred, it did not actually matter whether the underwriters additionally accepted that it had started accidentally since unless they alleged that it had been started deliberately with the connivance of the insured acceptance that a fire had occurred amounted to admission of the operation of an insured peril – see Slattery v. Mance

[1962] 1 QB 676

. However, somewhat to my surprise Mr Berry went further. On underwriters' behalf he accepted that the fire was accidental. I was surprised by this concession not least because it was not in any way foreshadowed in Mr Berry's written closing submissions which spoke in terms of the assumption that the fire was accidental. As will appear hereafter, in my judgment this concession does in fact have an impact upon the shape of the enquiry. It has given me cause for the most profound reflection as to the manner in which it should influence my evaluation of the evidence and my conclusion as to the probabilities. However I will revert to that later in this judgment. Underwriters denied that the insured could recover for a CTL. As to the actual total loss they asserted that a fire in the engine room and consequent explosions would not have caused and did not cause the flooding of hold No.4 or any double bottom tanks or any other spaces of the vessel than the engine room, or the sinking of the vessel. They had no positive case as to how seawater, at any rate in significant quantities, entered the engine room. They did however have a limited positive case to the following effect:- "…… the fire cannot have been the or a proximate cause of the sinking because there was so much water in the vessel from an unknown cause at the time the vessel was abandoned at about 1700 hours as cannot have been caused by an engine room fire. In particular there was about 2,700 metric tonnes of water in the vessel, of which about 1,950 metric tonnes was forward of Frame 36. It is to be inferred that water entering from this unknown cause both before and after abandoning was a proximate cause of the sinking." Finally, the Claimants for their part suggested that even if there were quantities of water forward of Frame 36 the reason for the ingress of which remained unidentified, nonetheless, unidentified ingress of water not being an excepted peril, they were entitled to succeed in their claim for an actual loss on the basis of concurrent causes – ingress of water into the engine room caused by fire which together with water which was forward of Frame 36 and had entered by unidentified means caused the loss of the vessel.

[6]The underwriters' case as to the quantity of water in the vessel and its location derives from various observations as to the vessel's draft and trim as to the reliability of which there was a lively debate at trial. The underwriters put forward no case at all as to how the water found its way into these spaces – nor even as to where it had come from. At least it can be said that it must have been seawater. There was unchallenged evidence – indeed it was in part the underwriters' own evidence – that the bilges were dry when checked at between 0800 and 0900 on the day of the fire. Furthermore, it was also the underwriters' own evidence that ballasting the vessel's double bottom tanks, which have a combined capacity of about 1,364 tonnes, took about five hours or so if done by gravity. By pump it would be quicker, although the only evidence as to the capacity of the ballast pump was that it was around 300 cubic metres per hour – maybe slightly more. For salt water a cubic metre needs to be multiplied by a factor of 1.025 to obtain the equivalent in tonnes. It was never explored at trial what might be the mechanism whereby 2,700 tonnes of water, 1,950 tonnes of it forward of Frame 36, could within the required time scale be introduced into the vessel after the bilges had been sounded and found to be dry. I mention the evidence at this stage only in order to demonstrate that if the unspoken suggestion is that the water was deliberately introduced I am not at all clear how or over what time scale that might have been achieved, consistently with evidence which is not simply unchallenged but in relevant part was adduced by the underwriters themselves.

[7]Although the underwriters challenged the reliability of some evidence as to the method whereby certain crew members entered the lifeboats, they did not challenge the bona fides of any of the factual witnesses. I have thus had no opportunity to see how these witnesses might have reacted to an allegation of misconduct had such an allegation been made. By the same token the shape of the trial resulted in my learning nothing of the owners beyond their corporate name. I do not know who is or are the true beneficial owner or owners and I know nothing of their financial circumstances. I do not know what was the actual as opposed to insured value of the vessel. I do not know if the vessel or other vessels in the same or associated ownership were trading at a profit or at a loss, nor whether there were difficulties in loan servicing. In any exercise of deciding whether the loss of a vessel can probably be attributed to a given cause, it is inevitable and I believe necessary that the court should at the very least consider the likelihood of alternative possible causes. Here no possible alternative cause was either advanced or even canvassed. The alternative left dangling in the air would involve criminal conduct instigated by persons whose identity I do not know carried out by persons who appeared before me and who were accepted to be witnesses of truth. I am not aware of any other case in which underwriters have admitted the occurrence of a fortuity which it is accepted could have caused the loss of the vessel, albeit that would be an unexpected outcome, and yet have put the owners to proof that that fortuity did cause the loss of the vessel without advancing any case as to the manner in which the vessel was in fact lost or even suggesting possibilities, inconsistent with the admitted fortuity having caused the loss. It has to be said that pursuit of such a strategy by underwriters and owners' response to it made for a trial which in many respects lacked reality.

[8]The Claim for a CTL . If the vessel had not sunk it would be axiomatic that the Claimants could recover for her CTL. The premise therefore upon which the issue arises whether the claim for a CTL is not maintainable is that the vessel sank other than by reason of the fire, or to put it another way, that the fire and the sinking are unconnected coincidences. This may of course be a false premise, but since it is the premise most favourable to underwriters it is not just necessary and logical to adopt it but it is also reasonable so to do. If the fire caused the sinking the debate is unnecessary.

[9]At first blush it is not easy to see why the claim for a CTL should not succeed. As of the moment when the damage to the vessel caused by fire reached the point at which cost of repair would exceed its insured value, the Claimants had an accrued cause of action against underwriters. Underwriters contend that that cause of action cannot be pursued in the absence of service of notice of abandonment, and that notice of abandonment cannot effectively be given once the vessel has become an actual total loss. Service of notice of abandonment is not however an essential ingredient of the cause of action, rather it is the notification of an election between two alternative remedies, payment for a partial loss or payment for a total loss: see Robertson v. Nomikos

[1939] AC 371

at p.377 per Lord Atkin and at p.381 per Lord Wright; Chandris v. Argo Insurance Ltd [1963] 2 LLR 64 and The Kyriaki [1993] 1 LLR 137. Furthermore the nature of the cause of action is for damages for the insurers' failure to indemnify the assured against or to hold him harmless from the losses he has suffered as a result of the damage to the vessel – see per Lord Goff of Chieveley in The Fanti

[1991] 2 AC 1

at p.35.

[10]Why therefore should the insured's ability to recover be affected by anything which happens to the insured property after his cause of action has accrued? In PYSBE v. Beer [1946] 79 LLR 417 Atkinson J. regarded the decision of the House of Lords in Andersen v. Marten

[1908] AC 334

as authority for the proposition that "once you have got a CTL, the mere fact that it may ultimately become an actual loss because of some event that is not within the policy does not affect your right at all to claim as for a CTL." Andersen v. Marten was in fact the reverse case to the present – CTL by capture, which was an excepted peril, followed by actual loss by perils of the sea.

[11]In order for underwriters to defeat the claim for a CTL there would have to be identified a rule of law which has the effect of removing the accrued right to claim. Atkinson J. also held that Section 77(2) of the Marine Insurance Act 1906 would be of no application in such a situation. That Section provides:- "Where under the same policy, a partial loss, which has not been repaired or otherwise made good, is followed by a total loss, the assured can only recover in respect of the total loss: Provided that nothing in this Section shall affect the liability of the insurer under the suing and labouring clause." Atkinson J. held that that sub-section is of no application where the original loss is a CTL. It may be that the common law rule is wider than the sub-section. Thus in British and Foreign Insurance Co. Ltd . v. Wilson Shipping Co. Ltd .

[1921] 1 AC 188

it was held that when a vessel insured against perils of the sea is damaged by one of the risks covered by the policy and before that damage is repaired she is lost, during the currency of the policy, by a risk which is not covered by the policy, in fact in that case by an excluded risk, the insurer is not liable for the unrepaired damage. However it should be remembered that both at common law and under the 1906 Act the time to estimate the loss where the party is put to no expense, as in the case of unrepaired damage, is at the expiration of the risk - see Livie v. Janson [1810] 12 East 648 and Marine Insurance Act 1906 S.69(3). See also per Lord Birkenhead L.C. in Wilson at pp.194-195. Wilson is no authority for the proposition that an accrued cause of action can be defeated by a subsequent loss.

[12]Neither authority nor reason would seem to support the proposition that where a CTL caused by an insured peril of which loss the insured is unaware is followed before he acquires such knowledge by a total loss caused by an uninsured peril or even by an excepted peril the insured cannot recover for the CTL. The problem of course only arises in what must be the very rare case where the insured peril which causes the CTL cannot be demonstrated to have played a causative role in the sinking, which is the premise upon which I am considering it.

[13]Section 61 of the Marine Insurance Act 1906 provides as follows:- "Where there is a constructive total loss the assured may either treat the loss as a partial loss, or abandon the subject-matter insured to the insurer and treat the loss as if it were an actual total loss." Mr Berry's submission was that once a vessel has become an actual total loss it is too late for an insured to exercise this right of choice in respect of a CTL which preceded the actual total loss but in respect of which the insured had not before the occurrence of the actual loss communicated an intention to treat the CTL as if it were an actual total loss. It is certainly the case that an insured who wishes to claim for a CTL must be prepared to abandon his vessel to underwriters and thereby indicate his willingness to treat the loss " as if it were an actual total loss" - see per Rix J. in Royal Boskalis v. Mountain, judgment delivered 18 December 1995 at p. 31 of the transcript. Abandonment is:- "A cession or transfer of the ship to the underwriter, and of all his property and interest in it with all the claims that may arise from its ownership, and all the profits that may arise from it, including the freight then being earned. Its operation is as effectually to transfer the property of the ship to the underwriter as a sale for valuable consideration, so that of necessity it vests in the underwriter a chattel of more or less value, as the case may be." See per Martin B. in Rankin v. Potter [1873] LR 6 HL 83 @ 144. Rix J. also observed in the course of his judgment in the Royal Boskalis case , at p. 32:- "Normally, the assured must indicate his election to abandon by giving a notice of abandonment, which thus becomes a condition precedent to the right to claim a CTL (Section 62(1)), but exceptionally a notice of abandonment may be dispensed with (Section 62(7), (8), (9)). The underwriter is not obliged to accept the notice, and thus the abandonment, but if he does the abandonment is irrevocable (Section 62(6)). It seems to follow, as Atkinson J. held, [in PYSBE v. Beer] that an unaccepted notice of abandonment is revocable. An underwriter may pay the claim without accepting the abandonment: he is not obliged to take over a wreck, which may be a damnosa hereditas – see Allgemeine Versicherungs-Gesellschaft Helvetia v. Administrator of German Property

[1931] 1 KB 672

at 687/88 per Scrutton L.J. If, however, he wishes, he is "entitled" to take over the assured's interest in the property (Section 63(1)). That is in any event the right of any underwriter who pays for a total loss, whether or not there is abandonment (Section 79, which repeats the language found in Section 63(1) to the effect that the insurer is entitled "to take over the interest of the assured in whatever may remain of the subject matter")." 15. Section 62 of the Marine Insurance Act 1906 provides:- "(1) subject to the provisions of this section, where the assured elects to abandon the subject-matter insured to the insurer he must give notice of abandonment. If he fails to do so the loss can only be treated as a partial loss. (2) Notice of abandonment may be given in writing, or by word of mouth, or partly in writing and partly by word of mouth, and may be given in any terms which indicate the intention of the assured to abandon his assured interest in the subject-matter insured unconditionally to the insurer. (3) Notice of abandonment must be given with reasonable diligence after the receipt of reliable information of the loss, but where the information is of a doubtful character the assured is entitled to a reasonable time to make enquiry. (4) Where notice of abandonment is properly given, the rights of the assured are not prejudiced by the fact that the insurer refuses to accept the abandonment. (5) The acceptance of an abandonment may be either express or implied from the conduct of the insurer. The mere silence of the insurer after notice is not an acceptance. (6) Where notice of abandonment is accepted the abandonment is irrevocable. The acceptance of the notice conclusively admits liability for the loss and the sufficiency of the notice. (7) Notice of abandonment is unnecessary where at the time when the assured receives information of the loss there would be no possibility of benefit to the insurer if notice were given to him. (8) Notice of abandonment may be waived by the insurer. (9) Where an insurer has re-insured his risk, no notice of abandonment need be given by him." The question which arises here is whether a claim for a CTL may be maintained where there has simply been no opportunity to give notice of abandonment before the vessel has been overcome by another peril and become an actual total loss in consequence of the operation thereof.

[16]It is obvious that in a case such as that under consideration cession or transfer of the ship whilst the vessel is and remains a CTL is not just impossible but of no conceivable benefit to underwriters. If the underwriters subsequently pay the insured's claim for a total loss cession or transfer will in any event take place by operation of law. Should recovery for a CTL be contingent upon the insured having performed a condition precedent performance of which is in fact impossible in a case such as the present, ie, service of notice of abandonment whilst the vessel is still a CTL? In Kaltenbach v. Mackenzie [1878] 3 CPD 467 at 471-475, Brett L.J. described the origin of the necessity of giving a notice of abandonment and explained its function. I make no apology for citing a very long passage from his judgment since it is in my view essential to look behind the formalities and to enquire precisely why notice is required to be given and what should be the consequences of either a failure or an inability to do so. Brett L.J. began his judgment as follows:- "This case raises the questions of abandonment and notice of abandonment on a policy of marine insurance. Before I enter upon the merits of the present case I think it desirable to state my view of the law. I agree that there is a distinction between abandonment and notice of abandonment, and I concur in what has been said by Lord Blackburn, that abandonment is not peculiar to policies of marine insurance; abandonment is part of every contract of indemnity. Whenever, therefore, there is a contract of indemnity and a claim under it for an absolute indemnity, there must be an abandonment on the part of the person claiming indemnity of all his right in respect of that for which he receives indemnity. The doctrine of abandonment in cases of marine insurance arises where the assured claims for a total loss. There are two kinds of total loss; one which is called an actual total loss, another which in legal language is called a constructive total loss; but in both the assured claims as for a total loss. Abandonment, however, is applicable to the claim, whether it be for an actual total loss or for a constructive total loss. If there is anything to abandon, abandonment must take place; as, for instance, when the loss is an actual total loss, and that which remains of a ship is what has been called a congeries of planks, there must be an abandonment of the wreck. Or where goods have been totally lost, as in the case of Roux v. Salvador , but something has been produced by the loss, which would not be the goods themselves, if it were of any value at all, it must be abandoned. But that abandonment takes place at the time of the settlement of the claim; it need not take place before. With regard to the notice of abandonment, I am not aware that in any contract of indemnity, except in the case of contracts of marine insurance, a notice of abandonment is required. In the case of marine insurance where the loss is an actual total loss, no notice of abandonment is necessary; but in the case of a constructive total loss it is necessary, unless it be excused. How, then, did it arise that a notice of abandonment was imported into a contract of marine insurance? Some judges have said it is a necessary equity that the insurer, in the case of a constructive total loss, should have the option of being able to take such steps as he may think best for the preservation of the thing abandoned from further deterioration. I doubt if that is the origin of the necessity of giving a notice of abandonment. It seems to me to have been introduced into contracts of marine insurance – as many other stipulations have been introduced – by the consent of shipowner and underwriter, and so to have become part of the contract, and a condition precedent to the validity of a claim for a constructive total loss. The reason why it was introduced by the shipowner and underwriter is on account of the peculiarity of marine losses. These losses do not occur under the immediate notice of all the parties concerned. A loss may occur in any part of the world. It may occur under such circumstances that the underwriter can have no opportunity of ascertaining whether the information he received from the assured is correct or incorrect. The assured, if not present, would receive notice of the disaster from his agent, the master of the ship. The underwriter in general can receive no notice of what has occurred, unless from the assured, who is the owner of the ship or the owner of the goods, and there would therefore be great danger if the owner of a ship or of goods – that is the assured – might take any time that he pleased to consider whether he would claim as for a constructive total loss or not – there would be great danger that he would be taking time to consider what the state of the market might be, or many other circumstances, and would throw upon the underwriter a loss if the market were unfavourable, or take to himself the advantage if the market were favourable. These are the reasons why I think the assured and the underwriters came to the conclusion that it should be a part of the contract and a condition precedent that, where the claim is for a constructive total loss, there must be notice of abandonment, unless there were circumstances which excused it. "Notice of abandonment, therefore, being a part of the contract, questions arose as to the time when that notice should be given. The first question which arose was whether the notice must be given at the first moment that the assured heard of the loss, or at some subsequent period. It was, however, decided that it is not at the moment of the first hearing of the loss notice of abandonment must be given, but that the assured must have a reasonable time to ascertain the nature of the loss with which he is made acquainted; if he hears merely that his ship is damaged, that may not be enough to enable him to decide whether he ought to abandon or not; he must have certain and accurate information as to the nature of the damage. Now, sometimes the information which he receives discloses at once the imminent danger of the subject-matter of insurance becoming and continuing a total loss; as, for instance, if he hears his ship is captured in time of war, it must be obvious to everybody, unless the ship is re-captured, it would be a total loss; or if he hears that the ship is stranded, and her back is broken, although she retains her character as a ship, if he gets information upon which any reasonable man must conclude that there is very imminent danger of her being lost, the moment he gets that information he must immediately give notice of abandonment. The law that has been laid down is, that immediately the assured has reliable information of such damage to the subject-matter of insurance as that there is imminent danger of its becoming a total loss, then he must at once, unless there be some reason to the contrary, give notice of abandonment; but if the information which he first receives is not sufficient to enable him to say whether there is that imminent danger, then he has a reasonable time to acquire full information as to the state and nature of the damage done to the ship. "But then there arose another question. Ships, or goods, or the subject-matters of marine insurance, are liable to danger at various parts of the globe, where neither the assured nor the underwriter is present; and upon the emergency the master of the ship being there alone, must act. Now, under those circumstances, masters have often sold either ship or goods; and masters have had to consider whether they would sell the ship or goods even in cases where such ship or goods are not insured. The general rule with regard to the propriety of a master selling the ship or the goods, is that he has no right to sell either the ship or the goods without the consent of the owner, but if necessity arises the master becomes what is called, from the necessity of the thing, the agent to bind his owner by a sale, or to bind the owner of goods by a sale. Now, the rule I should say from the necessity of things, at all events from the justice of things, is this, that if the circumstances are such that any reasonable person having authority from the owner would sell, then the master is entitled to sell, although he has not such authority. The question, I think, as between the person to whom a master sells and the owner of the property, is whether the circumstances were those which would have caused a reasonable owner, had he been present, to sell. If that state of things exists, the master has authority to sell, and his act is binding upon the owner of the ship or goods. Where, therefore, there has been a constructive total loss of either ship or goods, circumstances may have arisen which would justify the master in selling, or they may not; there may be a constructive total loss without any sale, and there may also be a constructive total loss accompanied by a sale. If the first information which the assured,, not being present, has of the damage which has occurred to his ship, or being the owner of goods of the damage which has occurred to his goods, although they were not an actual total loss by reason of the perils of the seas, is accompanied also by information that the master has sold, and if the circumstances of that sale were justifiable, so that the property passed to the vendee, under those circumstances that is the time when, if at all, the assured would be bound to give notice of abandonment; but in others that doctrine seems to be questioned. In Rankin v. Potter the law was established that where at the time when the assured receives information which would otherwise oblige him to give notice of abandonment, at the same time he hears that the subject-matter of the insurance has been sold so as to pass the property away, inasmuch as there was nothing of the subject-matter of the insurance which he could abandon, notice of abandonment was not necessary. No doubt the reason given for this was that notice at that time and under such circumstances would be a mere idle ceremony; it could be of no use. That was the point decided in Rankin v. Potter . In those particular circumstances it was held that notice of abandonment need not be given because there was nothing to abandon. That in one sense is true; but if goods had been sold it is obvious there must be something to abandon, that is the proceeds of the sale; the money which is the proceeds of the sale, when the insurance is settled, is abandoned; but where there is nothing of the subject-matter of insurance to abandon, there is no ship to abandon, there are no materials of the ship to abandon, there are no goods to abandon, notice of abandonment under those circumstances was said to be futile. But Rankin v. Potter went no further; it did not decide – because the point was not raised – that if, at the time when the assured had to make up his mind and when otherwise he ought to abandon, there was no sale of the subject-matter of the insurance, the assured would be excused from giving notice of abandonment if he was able to shew that, had he given such notice, in the result it would have turned out to be of no use. It was argued before us that the necessary inference to be drawn from Rankin v. Potter was, although there had been no sale of the subject-matter of the insurance when information of the disaster was received by the assured, yet if he could shew that before any notice of abandonment could reach the underwrite and before the underwriters orders could reach the assured a sale could take place, so that had the assured given notice of abandonment such notice would have been of no use to the underwriter, the assured would be excused from giving it. That point, however, is not raised here, and therefore it becomes unnecessary to decide it. I am not prepared to say that if it could be shewn that the subject-matter of insurance, at the time when the assured has information upon which otherwise he would be bound to act, is in such a condition that it would absolutely perish and disappear, before notice could be received or any answer returned, that that might not excuse the assured from giving notice of abandonment, but I am prepared to say that nothing short of that would excuse him; and although I do not say what I have stated would excuse him, I am not prepared to say it would not; that is the limit to which I think the doctrine could be carried, and it seems to me that to go further than that would let in the danger to provide against which the doctrine of notice of abandonment was introduced into the contract and made a part of the contract". In a case like the present allowing the assured to recover for a CTL does not involve that the assured is unilaterally throwing onto underwriters the risk of market fluctuation, or taking to himself the advantage of a favourable market. None of the reasons which Brett L.J. thought had informed agreement of the condition precedent would lead to the conclusion that an insured in such circumstances ought not to be permitted to recover for a CTL, at any rate not simply on account of failure to serve notice of abandonment before the vessel becomes an actual total loss by operation of a peril other than that which has caused the CTL. Having held that "where the assured at the time he receives the information on which he is bound to make his election has the thing insured in his power or under his control, he is bound to give notice to the underwriters" Cotton L.J. in Kaltenbach at pp 481-2 went on to leave open the very point which arises in this case:- "I give no opinion on the question which arises when the state of the thing insured is such that before the communication could have reached the underwriters it must, so far as human probability goes, have ceased to be in specie."

[17]It seems to me that in a case like the present the proper analysis is that abandonment of the subject-matter insured will take place by operation of law when the underwriters settle the claim. Section 61 is thus satisfied. The insured has been deprived of his right of choice envisaged by Section 61. He has no option but to treat the vessel as a total loss. Section 62(7) says in terms that notice of abandonment is unnecessary where when the insured receives information of the loss there would be no possibility of benefit to the insurer if notice were given to him.

[18]Mr Berry submitted that that analysis is wrong because in a case such as the present the insured has suffered no indemnifiable loss in consequence of the operation of the fire and/or explosions the real cause of his loss being whatever caused the sinking, the actual total loss. That is of course a proposition of causation. I can see that in the case of partial damage short of a CTL left unrepaired there may be scope for such an argument, and perhaps it is assisted by the fact that Section 69(3) prescribes that the recoverable measure of indemnity in such a case is the depreciation as at the expiry of the policy. In any event there is the very considerable authority of Lord Ellenborough in Livie v. Janson that in such a case the insured may not recover for the earlier partial loss. However in Hahn v. Corbett [1824] 2 Bing 206 , Livie v. Janson was distinguished. In Hahn cargo, Manchester cotton goods, was insured against marine risks from London to Maracaibo "warranted free from capture and seizure." The vessel ran onto the sand off Maracaibo and became a CTL. Ninety-five per cent of the cargo was damaged and there was no means to carry it on. On the next day, months before intelligence of the loss reached London, the goods were condemned as prize by the occupying Spanish forces and unloaded. The question arose whether the goods had been lost by the insured peril, perils of the sea or rather by seizure, which was excepted. It was held that the goods were lost when the ship was lost. In modern parlance that was a CTL of the goods and interestingly at the end of the report there is a note that in another case which only differed in that the goods were undamaged the same judgment was given. The crew abandoned the ship, or so it would seem, and it was unlikely that the insured could either recover the goods or carry them to destination. The case is to be distinguished from Roux v. Salvador [1836] BingNC266 where the goods, hides, were in a process of putrefaction which would have been complete by the time of arrival at the port of destination, Bordeaux, such that on arrival they could not properly have been described as hides, indeed in that case in the Exchequer Chamber Lord Abinger C.B. expressly said, at page 281:- "In the case before us the jury have found that the hides were so far damaged by a peril of the sea, that they never could have arrived in the form of hides. By the process of fermentation and putrefaction, which had commenced, a total destruction of them before their arrival at the port of destination, became as inevitable as if they had been cast into the sea or consumed by fire. Their destruction not being consummated at the time they were taken out of the vessel, they became in that state a salvage for the benefit of the party who was to sustain the loss, and were accordingly sold; and the facts of the loss and the sale were made known at the same time to the assured. Neither he nor the underwriters could at that time exercise any control over them, or by any interference alter the consequences. It appears to us, therefore, that this was not the case of what has been called a constructive loss, but of an absolute total loss of the goods: they could never arrive; and, at the same moment when the intelligence of the loss arrived, all speculation was at an end." Returning to Hahn v. Corbet t, the argument that the seizure should be regarded as the proximate cause of the loss was roundly rejected by the court consisting of Best C.J. and Park and Burrough JJ. In the unusual case where the subject-matter of the insurance is constructively totally lost by reason of an insured peril before being actually totally lost by operation of some other unconnected cause, it does not seem to me sensible to conclude that the assured has suffered no loss by reason of the operation of the first peril.

[19]Mr Berry pointed out that the admission by underwriters of the occurrence of the CTL related only to the scintilla temporis before sinking, at which time he said the vessel was doomed to become an actual total loss so that it was unreal to speak of her becoming a CTL before she was in fact an actual total loss. It is true that in their Amended Particulars of Claim the insured do not plead precisely when before the sinking the vessel became a CTL. The Claimants' expert marine engineer, Mr Bevis, estimated the cost of repairs to damage caused by an uncontrolled fire burning for 15 hours or so as US$7.9M if repaired in Colombo, US$5.8M if repaired in the Far East. He made no allowance for salvage or contingency. Mr Todd had this report of Mr Bevis before preparing his own. Without condescending to detail he considered that the likely cost of repairing damage likely to have been sustained prior to sinking would have "appreciably exceeded US$3M." On this state of the evidence and bearing in mind the hypothesis that the fire was not causative of the sinking I do not believe that I am justified in concluding that the vessel was already an actual total loss or bound to become one at the moment at which she became a CTL. Some unexplained event occurring after that point may well have tipped the balance between sinking and not sinking. I do not believe that my conclusion in this regard on the particular facts of this case throws up the spectre of owners being able, in most cases of actual total loss, to claim years after the event as for a CTL occurring a moment earlier than the actual loss. The problem only arises in the very unusual case where underwriters do not admit that the fortuity which caused the CTL caused the actual loss. In a case where actual total loss immediately follows CTL, such a situation must be very rare. A more realistic example would be a vessel which has become a CTL by fire thereafter being struck by an Exocet missile which causes her to sink. I cannot see why in such circumstances the insured should not recover from their marine risks underwriters. Of course those marine risks underwriters may have a subrogated claim against war risks underwriters if such there are, but that does not affect the analysis. The fact is that whereas the Exocet example is demonstrably two casualties and two losses, so the present case less obviously but for the purposes of the argument must likewise be regarded as two casualties and two losses. The premise upon which I am considering the claim for a CTL is that the actual total loss was brought about by a coincidental but unconnected cause.

[20]Mr Berry also placed reliance upon a passage in Arnould, Marine Insurance 16 th Edition at paragraph 1180 from which it appears that the distinguished editors of that work contemplate that some process of merger, analogous to that enacted in Section 77(2) of the Act, may take place in circumstances such as those under consideration here. Section 77 of the Act provides:- "(1) Unless the policy otherwise provides, and subject to the provisions of this Act, the insurer is liable for successive losses, even though the total amount of such losses may exceed the sum insured. "(2) Where under the same policy, a partial loss, which has not been repaired or otherwise made good, is followed by a total loss, the assured can only recover in respect of the total loss: Provided nothing in this section shall affect the liability of the insurer under the suing and labouring clause." The relevant passage in Arnould reads:- "1180. As we have already seen, an unrepaired partial loss merges with a subsequent total loss happening as the result of a later casualty occurring during the currency of the same policy, the assured may claim only in respect of that casualty, having no claim in respect of the earlier partial loss, whereas if the second casualty is due to the operation of a peril not insured against, the assured has no claim of any kind under the policy. "Although the point has never been decided, there are strong grounds for asserting that the rule is different in cases where there are successive total losses, each caused by separate casualties. The point cannot, of course, arise where the first loss amounts to an actual total loss. However there is no reason in principle why a loss which is constructively total should not be succeeded by a fresh casualty which can itself be treated as either giving rise to a constructive or to an actual total loss; although no doubt in the majority of cases the subsequent changes in circumstances would be held on the facts to flow from the original casualty. The propositions which follow are concerned with true cases of successive loss; cases where the ultimate loss is proximately caused by the original casualty involve no problem in the present connection. "In those cases, where there are, in the true sense, successive total losses, the following rules may be proposed. First, where the original casualty is covered by insurance and the assured has given notice of abandonment in due time, any casualty which may happen thereafter is irrelevant, so long as the original notice of abandonment is being maintained. Equally, if the abandonment has been accepted, a subsequent casualty cannot affect the issue. Thus if the second casualty is due to an excepted peril, the rule of merger which applied in the case of an unrepaired partial loss to defeat the claim has no application. On the other hand, it does seem clear that the assured, having properly claimed for a total loss, cannot, while maintaining that claim, make a second claim for another loss, either under the same, or any other policy. It is submitted, however, that so long as the first abandonment has not been accepted or the claim paid, it is open to the assured when a second casualty occurs, to withdraw his original notice and to claim in respect of the second loss under the same or another policy covering that loss, and he may well choose to do so if the second casualty appears to form a more secure foundation for a claim and if any doubt attaches to the validity of his initial abandonment. "Secondly, if the first casualty was not covered by insurance, or if notice of abandonment was not given in due time, the fact that the operation of the first peril produced a situation of constructive total loss, although not recoverable as such, does not preclude a claim from being made in respect of the second casualty, assuming it to be caused by an insured peril. "Thirdly, it is submitted that where the first casualty is treated as a partial loss, the right to recover in respect of that casualty is governed by the rules prevailing in cases where a partial loss is succeeded by a total loss. In other words, if the loss which was in fact constructively total, but not claimed as such, is left unrepaired, the loss merges with a subsequent total loss by perils operating during the currency of the policy, but if the second casualty is caused by perils operating after the policy has run off, no merger takes place." I should also set out footnote 80 which appears at the end of the first sentence of the last sub-paragraph of paragraph 1180. That reads:- "There appears to be no case where this point has been expressly decided, and it is arguable that since s.61 of the Marine Insurance Act 1906 merely confers a right of election on the assured as to the nature of his claim, not as to the nature of the loss (which remains a constructive total loss whether or not it can be claimed as such: see para 1170 supra ), the reference to "partial loss" in s. 77(2) must be construed as a reference to a loss which is in fact partial, not to a loss which is constructively total although not claimable as such. It is submitted that this view would lead to an absurd result, and that it is plain that whether or not the case is strictly comprehended within the wording of s. 77, the assured cannot elect to treat a first loss as partial, when it was in fact total, and then without repairing claim both for that loss and for a subsequent total loss occurring as the result of perils operating while the policy is still current. If , which is by no means clear, the dicta of Atkinson J. in P.Y.S.B.E . v. Beer , supra, at pp. 434 and 436 are to be read as suggesting that successive losses can be claimed under such conditions, then it is submitted that they cannot be supported. Brett L.J. clearly seems to have assumed in Pitman v. Universal Mar. Ins. Co .

[1882] 9 QBD 192

at p.208 that merger would operate under such circumstances, and it is submitted that this is the sounder view." I respectfully agree with the first sentence of the footnote. As to the second sentence, the example posed is indeed absurd but it is not of course this case. Subject to Mr Berry's independent argument on election arising out of the assured having claimed initially for an actual total loss, the assured here have in no sense elected to treat the first loss as partial nor have they attempted to claim for two losses. The footnote perhaps indicates also that what the learned editors had in mind, as indeed the text of the last sub-paragraph of paragraph 1180 seems to suggest, is the case in which it can really be said that the damage was left unrepaired as a result of some conscious choice. In my view if these distinguished and learned editors had intended to say that in a case such as the present the ability to claim for a CTL is lost in consequence of the inability to serve notice of abandonment, they would have said so. Mr Berry's submission involved that had the vessel sunk a little less quickly, and had the Master managed to make contact with his owners enabling them to give notice of abandonment at a time when the vessel was and remained a CTL, then underwriters would have had no answer to the claim for the CTL. On the footing that I am here dealing with two distinct casualties I do not believe that substantive rights should depend on such irrelevant fortuities as the owner having the opportunity, and I would add the requisite knowledge of the law of marine insurance, to give notice of abandonment.

[21]As already foreshadowed Mr Berry also submitted that by initially claiming for an actual total loss and maintaining that claim until their later amendment the assured necessarily asserted that the vessel was their property when it sank, a stance which is inconsistent with the abandonment or cession which is required under Section 61. That meant that the assured had elected to treat the earlier loss as partial because they had elected to keep the property until it sank. It was not suggested that this was any species of estoppel – it was said to be an irrevocable election. The short answer to this point is provided by Rix J. in the Royal Boskalis case at page 32 where he regarded it as implicit in the reasoning of Atkinson J. in PYSBE v. Beer that a notice of abandonment, if not accepted, may be revoked, expressly or by conduct. At page 33 of his judgment Rix J. said this:- " It is common ground that there is no pleaded case of equitable estoppel, and that if, therefore, the sole basis for the doctrine of revocation of an abandonment is that of equitable estoppel, then the defendants' point is a bad one. It seems to me, however, that if an assured, whether before or after commencement of his action, expressly revokes his notice of abandonment, or declares that he will make no claim for a total loss in circumstances where he has recovered his property, then he has chosen or elected to forego his CTL claim and to retain his property. I do not see why he should not be entitled to do so, nor why he should not be bound by that. If he can do it expressly, he can do it impliedly, or by conduct." If notice of abandonment is revocable, so equally must the "election" constituted by claiming for an actual total loss be revocable. The longer answer to Mr Berry's submission, although perhaps in fact I can state it more shortly, is that in any event I do not accept that by commencing and maintaining an action for an actual total loss the assured either communicated an unequivocal statement that an alternative or logically prior claim for a CTL would not thereafter be advanced or conducted themselves inconsistently with a willingness to cede the insured property to underwriters. Indeed it was implicit in what the assured did that the insured property would be ceded to underwriters on payment of their claim, as it must be – see Section 79 of the Act. In this regard there is no difference between an actual and a constructive total loss. In the circumstances of this case with the vessel lying on the seabed (in very deep water as the underwriters would have wished to remind me had they had any evidence that that was so) no one could have thought that the assured were asserting a proprietary interest adverse to underwriters.

[22]Finally in this regard Mr Berry relied upon Clause 18.2 of the Institute Time Clauses Hulls which provides, under the rubric "UNREPAIRED DAMAGE":- "18.2 In no case shall the Underwriters be liable for unrepaired damage in the event of a subsequent total loss (whether or not covered under this insurance) sustained during the period covered by this insurance or any extension thereof". I do not consider that this clause avails Mr Berry. If the assured are right this far they are not seeking to render the underwriters liable for unrepaired damage – they are seeking to treat the loss as if it were an actual total loss. This clause adds nothing to the debate.

[23]For all these reasons I conclude that the Claimants are in fact entitled to recover as for a CTL. Notwithstanding that conclusion I must of course go on to consider what would be my conclusion if I had to consider the alternative claim for an actual total loss.

[24]The claim for an actual total loss. It will already be apparent from my brief recital of the circumstances of the casualty in paragraph 2 above that there was virtually no first hand evidence in the shape of first hand observations of the nature, scale or spread of the fire. The engine room was so swiftly abandoned that few first hand observations could be made. I heard oral evidence from seven ship's witnesses. The owners called to give evidence the Master, Mr Konizaris; the Chief Engineer, Mr Bougiduris; the Third Engineer, Mr Belsky; the Second Officer, Mr Saveljeus; the Bosun, Mr Kostelis; and an oiler, Mr Ngarawa. The underwriters called the Chief Officer, Mr Borisovs. The Master, Chief Engineer and Bosun were Greek and gave evidence in the Greek language through an interpreter. The Third Engineer was from the Ukraine, the Chief Officer and Second Officer were from Latvia: all three gave evidence in the Russian language through an interpreter. The oiler was Tanzanian and gave evidence in Swahili through an interpreter. Notwithstanding it was given through an interpreter their evidence did not take very long – it occupied only about three days of court time. Additionally the owners put in under the Civil Evidence Act 1995 hearsay evidence in the shape of witness statements from a fitter, Mr Onyeko (Ukrainian); AB Seaman, Mr Mwakidole (Tanzanian); Oiler, Mr Marinyimya (Tanzanian); Oiler, Mr Kasudi (Tanzanian); Second Engineer Mr Smettannikous (Latvian); Electrician, Mr Babajav (Lithuanian); AB Seaman, Mr Elefu (Tanzanian); General Purpose Crewman, Mr Okiula (Tanzanian); General Purpose Crewman, Mr Mwaroko (Tanzanian); Chief Cook, Mr Macoha (Latvian); Assistant Cook, Mr Arturas (Lithuanian) and a Mr Petrongonas, an independent marine technical consultant practising in Athens who travelled to Colombo to attend the interviews of crew members. The crew were rescued by a German vessel after the sinking and taken to Sri Lanka. They were disembarked off a southern port and travelled from there to Colombo. Interviews were conducted jointly by Mr John Dillon of Messrs Ince & Co on behalf of owners, and Mr Bill Kerr of Messrs Holman, Fenwick & Willan for underwriters. Mr Joe Quain of Messrs Bentleys, Stokes and Lowless was present on the second and subsequent days of interviews on behalf of the Owners' P and I. Club. A Mr Stankevics of the Owners' crew manning agent was also present. He assisted where necessary with translation from Russian, the first language of at any rate those witnesses from the countries within the former USSR who gave evidence before me. There was present no person capable of translating between Swahili and English. According to Mr Dillon the Tanzanian crew members spoke good English. Although there were some supplementary statements prepared in a different manner for the most part the first and in most cases the only statement of each witness was prepared by Mr Dillon on his return to his office in Singapore on the basis of his detailed notes of the interviews. Mr Dillon has experience at sea as a deck officer and is enormously experienced in interviewing crew following a casualty and in preparing statements for use in litigation. I believe that Mr Kerr has a similar background. Unusually in my experience the court had available to it transcripts of tape recordings of most of the interviews, as well as the solicitors' extensive notes. This meant that it was possible at trial to be critical of certain occasions when more or less important observations appear to have been obtained in response to leading questioning. Having said that I do not underestimate the difficulty of extracting a coherent account in such circumstances. Whatever the cause of the loss, the crew members had been through a disorientating experience, about which they were now being asked questions, not in their mother tongue, by lawyers representing different interests under the gaze of their employers to whom they owed their livelihoods. In the light of observations during the trial as to the nature of this exercise, including observations by me, I feel that it is important to place on record that I have no doubt whatever that the solicitors on both (or all) sides did their conscientious best accurately to extract and record the honest recollections of the witnesses and to probe that evidence so far as could usefully be achieved within the confines of the exercise. It is not to be wondered at that the intensity of the spotlight thrown at trial upon one or two aspects of the evidence, particularly so far as concerns subjective evaluations of the strength of such explosions as may have occurred, revealed areas in which answers could not necessarily be relied upon, at any rate at full face value. Furthermore, it should be borne in mind that those conducting the questioning cannot have foreseen the importance which might in due course be attached to seemingly unimportant matters – such as the precise location of the pilot station at which some crew members disembarked. Two possibly related weaknesses in the procedure were however thrown up at the trial. The first is that the prominence afforded to an event such as the possible explosions was sometimes exaggerated in the draft statement offered to the witnesses for signature. Secondly, the statements did not save in one case record how the witness had before signing the statement satisfied himself that the statement given therein in the English language was in all respects accurate. Indeed the Bosun in his oral evidence stated that he had had no translation of what he signed and that he took no steps to satisfy himself that what he signed on two separate occasions corresponded with what he had said to those who interviewed him. This serves to emphasise, if it needed to be emphasised, the reserve with which evidence in chief must now be approached in a system in which witness statements almost invariably stand as such evidence.

[25]I have naturally placed most weight upon evidence which was explored or tested at trial and which was not fundamentally challenged as to its accuracy or veracity, or is incapable of challenge by the party adducing it. This latter observation principally but not exclusively relates to the Chief Officer's evidence adduced by underwriters. Mr Berry rightly reminded me that the account which I was invited by the Claimants to adopt was in some respects at variance with evidence which the Claimants had adduced as hearsay under the Civil Evidence Act. For example, many of the witnesses whose evidence was given by hearsay statement do not state that they heard loud explosions, whilst some spoke positively that they heard no explosions, loud or otherwise. Mr Berry also pointed out that an inflexible application of the rule that a party may not challenge the accuracy of his own evidence would in this case lead to some difficulties. For example, the Chief Officer said that what came to be called the second explosion was loud whereas the Chief Engineer said that there was no such explosion at that stage. I would agree that the rules must be applied with commonsense, without slavish adherence to form. With those observations in mind I am able to make the following main findings of fact.

[26]The vessel departed Aqaba with a cargo of 17,400 tonnes of rock phosphate loaded into all four holds. Her departure draft was 9.3 metres forward and 9.86 metres aft. Normal consumption would tend over time to bring her towards even keel. The weather on 9 March was good with only a slight sea and a small swell. The bilges were dry when checked between 0800 and 0900. The Third Engineer came on watch in the engine room at 1200. Between 1215 and 1230 he carried out a detailed round of inspection of the engine room and found nothing amiss. At about 1330 he did another inspection of the engine room that took about 15 to 20 minutes. Nothing was amiss. Fuel oil Purifier No.2 was running correctly. That was an item which had received special attention by way of inspection only something of the order of thirty minutes before fire was noticed in the engine room. Oiler Ngarawa was on duty in the engine room at this time. About half an hour before the discovery of the fire he checked the purifier area for, amongst other things, oil leaks. He found no leaks. About 15 to 20 minutes before the discovery of the fire he went to the boiler flat and, after topping up the water tank, began to clean soot from the boiler tubes using a steam lance.

[27]First observation of the fire was at about 1420 hours on 9 March. At that time the Chief Engineer and Third Engineer were in the engine room control room. The Chief Engineer had come to check quantities of fuel remaining on board. Oiler Ngarawa was by now on the boiler flat. No one else was in the engine room at the time and no one else saw the fire apart from a fleeting observation by the electrician through an access door to the engine room. The Chief Engineer and the Third Engineer saw fire and smoke in the purifier area generally, near the gasoil service and settling tanks which are forward of the purifiers on the starboard side. There was more smoke than flame. They were not able to see from where the fire originated. The Chief Engineer stopped the main engine and told the Third Engineer to sound the general alarm, which caused Oiler Ngarawa to attempt to return to the engine room control room to see what the problem was. When he had reached about halfway down the ladder connecting the boiler flat to the upper machinery flat he saw black smoke coming from the purifier area. He saw predominantly smoke, with perhaps some flashes of flame in it, although he thought that the fire was already too big to be controlled with hand held extinguishers. The Third Engineer took a 5 kg hand held CO2 extinguisher from the engine room control room towards the area of the fire. Together with the Chief Engineer he approached the purifier area but could get no closer than the stairway on the starboard side because of the smoke and fire although it was smoke which made more impression on them than flames. The Chief Engineer told the Oiler to leave the engine room quickly. The Chief Engineer and the Third Engineer also left the engine room within about two or three minutes of their first observation of the fire. The Chief Engineer asked the Third Engineer to go and open the valves on the fire pump at bottom plate level. The Third Engineer suggested that it was too dangerous to do so and that there was likely to be fire lower down in the engine room. The Chief Engineer therefore instructed the Third Engineer to stop the engine room ventilation fans, the switches to which are located in the engineers' changing room on the port side of the boiler flat, which he did, going out into the port alleyway. On leaving the engine room the Chief Engineer operated the pneumatic trip for the quick closing fuel tank valves. However at least one of the generators, two of which were on load, continued to run for about an hour suggesting that some or all of the valves did not close.

[28]The Second Officer was on watch at the time of the general alarm. He had neither seen nor heard anything unusual prior to that time. The Master was woken by the sound of the alarm and was on the bridge within two or three minutes. The Chief Officer was also asleep when the alarm was sounded.

[29]On making his way out of the accommodation the Chief Engineer met the Second Engineer whom he instructed to close the engine room skylight and ventilators and the funnel dampers and he met also the electrician whom he instructed to shut off the ventilation fans in the galley and the air conditioning in the accommodation. Appropriate preparations were made for fire fighting but none was attempted. Smoke could be seen coming continuously from the funnel notwithstanding closing of the engine room ventilators and skylight and the funnel damper.

[30]The Bosun put on the fire suit with the intention of entering the engine room. The plan was that he should reach the large foam extinguishers mounted on a trolley, somewhat inappropriately called the foam chariot, which is located near to the fuel oil purifiers. However the volume of smoke being produced, which was now arriving also at the alleyway at main deck level, at the entrance to the engine room, persuaded the Chief Engineer that the Bosun could not enter the engine room and in consultation with the Master he decided to use the fixed CO2 smothering system. That decision may also have been influenced in part by noises coming from within the engine room which some witnesses described as explosions, although they did not speak with one voice. For example, I do not believe that the Third Engineer's evidence, evaluated as a whole, supported the occurrence of a significant explosion at this early stage. However, the Chief Engineer, who I regarded as giving me the most reliable assistance in this area, did not regard any explosions at this stage as of any significance and since the Master in his evidence did not mention any explosion at all at this early stage in the sequence I conclude that the occurrence of one or more explosions did not inform the decision to use CO2 and that such explosions as had by this stage occurred cannot have been significant.

[31]The Chief Engineer's first attempt to get to the CO2 room was thwarted by smoke in the cross alleyway on the main deck level. He had to return to the poop deck. He considered putting on breathing apparatus but, rather than waste time, put a wet towel over his head and made a second attempt. When he reached the CO2 room he fired the pilot bottles which discharged the contents of the CO2 bottles into the engine room. In the meantime other members of the crew were using the fire hoses, powered by the emergency fire pump, to carry out boundary cooling of the decks and accommodation. The accommodation bulkhead was sufficately hot to cause some of the water sprayed onto it to turn to steam. Immediately following the release of the CO2 there was a noticeable drop in the level of smoke coming from the funnel area. It is impossible to say when precisely the CO2 was released – it is unlikely to have been before about 1500 hours and could have been up to about 30 minutes later.

[32]About 30 minutes after the CO2 had been released the smoke escaping from the funnel area and starboard engine room ventilator began to increase significantly, which caused the Master, following a discussion with the Chief Engineer, to instruct the Chief Officer and the Second Officer to prepare the lifeboats, load them with provisions and stand by. Again there was a considerable body of evidence to the effect that the lack of success of the CO2 was heralded by a loud explosion followed by an increase in the volume of smoke being emitted. However I think that this is unlikely since the Master accepted in cross examination that there was no "very big" explosion at this time. The Chief Engineer's evidence in cross-examination, consistent with an emphatic correction which he made in interview in Colombo, was that it was the further smoke which demonstrated that the CO2 had not been successful. It is clear from the evidence of the Master and the Chief Engineer that the decision to prepare the lifeboats was not informed by the severity of any explosion which may by then have occurred. If there was an explosion 30 minutes or so after the CO2 was released it was not of significant proportions. I have not of course in this regard overlooked the Chief Officer's evidence. However that evidence was not of high quality and does not in my judgment justify a finding that there was at this time an explosion of significance, such that it informed the decisions thereafter taken. In his interview the Chief Officer had said that "at this moment I listen not very heavy, but explosion" which he said he could feel through his feet. In his first witness statement this had become a large explosion although in his second witness statement he said "the only explosions I heard or felt …. were like a "thump". None were strong enough to cause the ship to shudder or shake." In cross examination there was put to him the Master's witness statement to the effect that there was a loud explosion, not his answer in cross examination to the effect that there was no very big explosion at that time. The Chief Officer, who had recently obtained further and much sought-after employment with the owners, tended to agree with such comments of the Master as were put to him, and he agreed with this one. I have already said that I have found the Chief Engineer's evidence on the severity of explosions of the most assistance, and I cannot attribute more weight to this response to a leading question than I do to the Chief Engineer's evidence, corroborated as it is by the Master's evidence, as to what it was that prompted preparation of the lifeboats. It is fair to say that the Third Engineer's evidence probably does support an explosion at this stage – indeed he changed his manuscript statement to make reference to a large explosion at this stage rather than the "audible noise" to which the first draft referred. However the Third Engineer did not remember the sequence and number of explosions and of them all he remarked "maybe for somebody else it could be perceived as just noise". Again I do not think that this evidence should persuade me to a finding that it was a loud explosion which indicated the lack of success of the CO2 and prompted the decision to prepare the lifeboats when that is not supported by a fair evaluation of the evidence of the decision makers, the Master and the Chief Engineer, and where I have more confidence in the Chief Engineer's judgment and realism in evaluating explosions than that of the other witnesses.

[33]Contemporaneously with the preparation of the lifeboats observations were made that the temperature in and around the accommodation was increasing. Deck plates were becoming too hot to walk on. Paint was observed to be burning on both sides of the ship at engine room level. It was the increase in temperature which prompted the Master now to give the order to abandon ship and to send a distress signal. This he did by pressing the automatic VHF and telex alarms on the vessel's GMDSS system. Records for the Falmouth Maritime Rescue Co-Ordination Centre indicate that the distress alert was received at 13.49 hours UTC, which translates to 16.49 hours local time on board the vessel. All but the Master and Chief Engineer boarded the lifeboats, the Chief Officer in command of the port lifeboat, which had no motor. The Master ordered that the lifeboats be moved forward, away from the danger area immediately adjacent to the engine room. The Master returned to his cabin to collect passports, certificates, cash and some spare clothes.

[34]It was probably whilst only the Master and the Chief Engineer remained on board that there then occurred a loud explosion. If there had been earlier explosions this was of a greater order of magnitude and significance, however I do not feel justified on the totality of the evidence in finding that it caused significant vibration. The Master's evidence that "the whole ship shook" must I think be regarded as hyperbole in the absence of any comment on vibration from the Chief Engineer. The Chief Engineer appeared to me to give more careful thought to his evidence than did the Master. I am not convinced that any other person was actually on board when this explosion took place. The Chief Officer gave some graphic evidence of an explosion which he felt through his feet to be considerable, but this was at a time when he was in his cabin. It is possible that the Chief Officer was still on board at the time of this significant explosion, i.e. that which caused the Master and Chief Engineer rapidly to leave the ship, but if he was he was himself on the point of entering the port lifeboat, not in his cabin. In his second Witness Statement the Chief Officer had said, as I have already observed, that the only explosions which he heard or felt while he was on board the ship were like a thump, and that none was strong enough to cause the ship to shudder or shake.

[35]On the other hand it is clear that this significant explosion was accompanied by, or produced, a noticeable escape of smoke in the area at the after end of No. 4 hold, in front of the accommodation. I think it most unlikely that this was simply smoke escaping from the accommodation. This was smoke which was associated with the explosion, and the explosion was heard to come from the engine room, not from the accommodation. I think that it is most unlikely that there was any fire in the accommodation at this stage, and the arrangement of double doors connecting the accommodation with the deck space would in any event militate against any escape of smoke from the accommodation through the doors. The Chief Engineer was quite clear in his evidence that the smoke did not come from the accommodation. The Chief Engineer thought that the smoke had either come through the hatch which gives access to No. 4 hold or possibly from vent pipes to the double bottom fuel tanks which run aft of the engine room bulkhead but emerge onto the weather deck just forward of the bulkhead on either side of the access hatch. The Master's thoughts were less focused than those of the Chief Engineer but he clearly associated this explosion and its manifestation by smoke with the aft end of No. 4 hold, being specifically concerned that it might have affected the integrity of the bulkhead.

[36]The Chief Officer was called to give evidence by the underwriters in the hope and expectation that he would say that the vessel's midships draft on abandonment was 10.5 metres. He had said this in a supplementary Witness Statement taken by the underwriters' solicitors two months before trial, in consequence of which no doubt the Claimants announced that they would not after all call him as a witness as they had hitherto intended. In the event the Chief Officer resiled from this proposition even in his evidence in chief. Furthermore in his Supplementary Witness Statement the Chief Officer spelled out in terms that he had been asked by the Master to tell him when the draft became 10.5 metres and that when in due course he did do the Master responded by saying "we must abandon ship.'' The Chief Officer did not maintain this account in his oral evidence. The Chief Officer's Supplementary Statement was dated 29 th April 2002. It was in early May after 3 months of trying that the Chief Officer obtained the much sought-after employment to which I have already referred. He found himself in an invidious position. It was no doubt partly as a result of that that his evidence was not, as I have already remarked, of high quality. I found this a very unsatisfactory part of the case. I had already observed whilst the Master was giving evidence that he was very sensitive when being questioned about what the Chief Officer had said in his Supplementary Witness Statement, from which the Chief Officer had not of course by then resiled as I have now recorded. I am left in very real doubt whether I have received an entirely full and accurate account of all that passed between the Master and the Chief Officer on the topic of draft. In consequence my findings are very limited. I am satisfied that at the beginning of the episode some time after the General Alarm had sounded, the Master asked the Chief Officer to check the vessel's draft. The Chief Officer checked the midships draft at least twice and he attempted in between those two occasions to check the forward draft. The midships draft he checked by leaning over the rail of the port, lee, side and observing as best he could the relationship between the water level and the painted draft marks. The ship was rolling and there was swell, so the exercise involved estimation and judgment. At the first check the Chief Officer thought that the midships draft was normal, i.e. unchanged from what it should have been having regard to the Aqaba departure draft. He could not form any view as to the forward draft. The last occasion on which the Chief Officer checked midships draft was shortly prior to the order to the crew to abandon ship. On this occasion the Chief Officer estimated from his observation that there had been an appreciable increase in the midships draft. In his oral evidence he used the expression "a practical increase.'' By this he meant I think an increase of the order of about one metre, because he would have regarded an increase of that order as(a) reasonably observable and

(b) obviously significant. That is why he regarded 10.5 metres which was apparently stuck in his mind "as a milestone figure.'' If the increase were of the order of 1 metre it would mean that the draft had reached 10.5 metres. Obviously, since the increase in draft was estimated in difficult circumstances, it may well have been less than a metre. But having regard to the totality of the Chief Officer's evidence and his explanation that 10.5 had been mentioned by him because stuck in his mind "as a milestone figure" I think it most unlikely that the increase was less than of the order of half a metre. Below an increase of that order I doubt if the Chief Officer would have registered it as a practical increase. The Chief Officer told the Master that there had been an increase in draft. I think it is a fair inference that the Master would have interpreted that as meaning an appreciable increase. I do not have enormous confidence in the Master's denial that the information concerning draft was of any relevance whatever to his decision to order the crew to abandon ship. On the other hand I am satisfied that that order was principally motivated by the increase in temperature of accommodation and deck plates to which I have already referred.

[37]Some time was spent at trial exploring the manner in which the three senior officers – Master, Chief Engineer and Chief Officer, entered the lifeboats. The Claimants' purpose in this regard was to paint a picture inconsistent with a midships draft anything other than only a little larger than the pre-incident draft. I found this exercise singularly unhelpful and unconvincing. In his Witness Statement the Chief Engineer had said that he and the Master boarded the lifeboat via a pilot ladder, an account confirmed by the Bosun in his Witness Statement. The Bosun had been in the starboard lifeboat when the Master and Chief Engineer boarded it, as had the Second Officer and the Third Engineer. In oral evidence the precise method of boarding of the lifeboats was canvassed with the witnesses. The totality of their evidence on this issue, recorded at interview, enshrined in written statements and given orally at the trial was so confused and contradictory that I find it quite impossible to make any reliable findings, certainly none on the basis of which I could derive an inference as to the then midships draft. It was suggested that the reference to a pilot ladder had been a confusion or misunderstanding when the intended reference was to a pilot station. That might be thought an unlikely mistake for experienced seamen to make. The interviewers were seafarers too. But then there was complete confusion as to where the pilot station was. A photograph seemed to show the pilot station forward of midships, between Nos. 2 and 3 holds. The witnesses were adamant that it was further aft, between Nos. 3 and 4. I thought it significant that the Third Engineer declined an invitation put to him at the end of a very short examination in chief to describe the manner in which the Master and the Chief Engineer had entered the lifeboat. He prayed in aid seasickness, the energy which he was devoting to looking after the engine and shock as contributing to a physical and mental condition which would not allow him to observe who was boarding the lifeboat or what was the manner of their doing so. The Third Engineer was a careful and direct witness who struck me as being of at least usual fortitude. I inferred from his answer that it was an area about which he preferred not to comment. The Master and the Chief Engineer may well have scrambled into the lifeboat in the manner which they described, hanging from the rail at the side of the ship and being grabbed below the knees and hauled into the boat by those already in it, although obviously my confidence in that account is not assisted by the Chief Engineer's and the Bosun's earlier reference to use of a pilot ladder. This evidence alone would be a slender basis for drawing any inferences concerning draft. However when that uncertainty is combined with uncertainty as to the location at which they boarded, it becomes clear that the evidence will not bear the weight sought to be put upon it. Thus if the Master and the Chief Engineer boarded from a point at which there was no gap cut into the bulwark, so that they were hanging not from the bottom of upright rails each side of the gap but rather from the top horizontal rail of the bulwark, the draft would I think have to have been considerably increased to enable them to have been grabbed below the knees by those in the boats. Furthermore scale diagrams drawn by the naval architects showed that dangling from the bottom of upright rails with the vessel's draft at 10.5 metres was not inconsistent with being helped into the boat by being grabbed below the knees, since even on this assumption the dangling men's feet would have been well above the bottom of the lifeboat. However the evidence on boarding of the lifeboats generally was so unsatisfactory that I cannot safely make any findings which assist on the question of draft. I accept that the Chief Officer jumped into the port lifeboat, timing his jump carefully so that the effects of roll and swell combined to produce optimum conditions. I am not sure whether I can deduce anything from that as to the feasibility of what is suggested to be the method of boarding used by the Master and the Chief Engineer, and of itself it tells me nothing about draft. Any attempt to draw inferences would have to include an evaluation of the Chief Officer's courage, particularly under the exigencies of the moment. Likewise I cannot reach any reliable conclusions as to the extent to which those in the boats were able to see onto the main deck of the vessel. It is not at all clear to me that any of them would have been standing in the boats, let alone on the bulwark of the boats if that is possible and it is very clear to me that a precise evaluation of whether they could see onto the main deck of the vessel which they had just left would not have been uppermost amongst their priorities. None of this evidence would bear the weight sought to be put upon it.

[38]Once everyone was in the boats the starboard boat manoeuvred around the forward part of the vessel to take the port boat in tow. The two boats were then distanced from the vessel, probably ending up several hundred metres astern. Once outside the vessel the Chief Officer was able to observe that the trim by the stern had increased. It is impossible to say by how much the trim had increased. Perhaps the proper inference from the combination of the Chief Officer's evidence and the Bosun's evidence that the trim was unchanged is that the increase was noticeable but not great. From a distance of several hundred metres it was possible to observe that the Port of Registry letters on the transom were clearly visible and distinctly above the waterline. Some of those in the port lifeboat observed as they moved away from the vessel that paint was peeling form the port side of the vessel in way of the engine room. One of the witnesses described this as paint on the hull or above the waterline bubbling. From the boats it was possible to observe through the night that a fierce fire was burning in the accommodation. One large explosion that was heard was almost certainly oxyacetylene bottles stored on the main deck just forward of the main accommodation. Between 0500 and 0600 hours next morning the vessel sank stern first.

[39]The Expert Evidence It will be apparent from the foregoing recital of what reliably is known of the loss that there was very little solid material upon which the experts could base their opinions as to what might have occurred and, in particular, how the engine room fire could plausibly have lead to the sinking of the ship within the known timescale. This paucity of material did not act as any noticeable deterrent to speculation. Whereas the factual evidence, given through interpreters, took about three days the expert evidence occupied nine days. As is customary in this type of litigation most of the experts found themselves more or less politely accused of partiality, usually disguised as lack of objectivity. As best I can recall the naval architects were largely exempted from this accusation. Although not always easy to follow their evidence was very helpful, although at the end of the day their contribution to the case could only be small, albeit highly significant. Two important and central points emerged from their evidence. The first was that for the vessel to sink it was necessary for hold No. 4 to have been flooded in addition to the engine room. The second was that there were two distinct means whereby, broadly, No. 4 hold might become flooded. Firstly, water might in some manner be introduced into the hold below main deck level. Secondly however there was the possibility of downflooding into No. 4 hold caused as a result of an opening into No. 4 hold at main deck level becoming close to the static equilibrium water line. I put this last point deliberately imprecisely because there was a debate between the naval architects whether the opening actually required to be submerged below the static equilibrium waterline for there to be a real risk of downflooding. On this point I found the evidence of Mr Parry for the Claimants more persuasive than that of Mr Boyd for the Defendants. Of course if a vessel is perfectly still in perfectly calm water then downflooding ought not to occur through openings which are above the static equilibrium waterline. However that situation will rarely obtain at sea and it did not obtain here. There are some obvious uncertainties in any event. For example, the "opening" in question here is the access hatch to No. 4 hold which is contained within a closed area sometimes described as looking like a telephone box into which leads a door above a threshold. The lower edge of the door is 600mm above deck level. The hatch itself sits on a sill which extends 350mm above the deck level. The hatch was probably closed but I am not aware of any evidence as to its condition or as to whether it was secured. In any event it was designed only to be weathertight not watertight. Mr Boyd adopted a principled but I thought formulaic approach which involved a mechanistic acceptance of the philosophy underlying the relevant load line regulations to the effect that so long as openings remain above the waterline downflooding is not considered to occur. I am sure that that approach is entirely justified in ship design and in any event I have no doubt that the design approach to the requisite safety margin is much more sophisticated than I have crudely paraphrased. I also understand that the relevant rules are more concerned with side impact damage than with the effects of a prolonged engine room fire. However that may be, if one is considering retrospectively whether downflooding might have occurred it seems to me more realistic to proceed on the basis that progressive flooding could well occur if the remaining freeboard was 30cm or less, which was how Mr Parry put it. As I understood the evidence if the engine room floods, and if other compartments or tanks aft of frame 36 flood and/or one or more of the No. 4 double bottom or top side fuel oil tanks flood then the vessel was in the prevailing weather conditions at risk of sinking through downflooding into No. 4 hold. The disagreement was really only as to the magnitude of the risk. Mr Boyd did not rule out the possibility of sinking occurring in this manner. As I have indicated, I consider that the risk was appreciable, but this disagreement between the naval architects was of little moment since in my view there are formidable difficulties in the way of a conclusion that the fire might plausibly have brought about the necessary conditions within the timescale required. It might have done, and I need to consider whether in my overall evaluation of the probabilities I am compelled to the conclusion that it did. That however would be a conclusion which would involve an evaluation of the entirety of the case – certainly an exercise going beyond the ambit of the opinions which the expert witnesses were qualified and entitled to express.

[40]In that latter regard I believe that the fire and marine engineering experts on both sides were placed in a difficult position and I would not wish to be unduly critical of them. I believe that they saw themselves as engaged upon an exercise which was to be conducted according to restraints and limitations devised by the lawyers which they might not themselves have regarded as realistic or which might not have involved a sensible starting point for their enquiry. Thus the fire experts on both sides were asked to prepare their reports upon the assumption that the fire was accidental. Dr. Foster, the Claimant's fire expert told me that he had not "focused to any great degree" on how the fire might have started – because he took it as a given that there had been a fire at a certain stage and he had not been asked to consider in detail how that fire started. This was I think a serious shortcoming in his evidence. It is obviously of the first importance, I should have thought, in assessing the likely development of the fire and in particular the speed at which it may have developed to consider how much fuel or other flammable material might have accumulated by the time of ignition and to what extent the source of that fuel represented a continuing source from which further fuel could be derived to sustain and augment the fire and to enable it to spread. There seems to be general agreement that the most likely source of fuel in the first instance is a leak from a high pressure line or valve associated with the IFO heaters which are close to the shell plating on the starboard side, forward of the IFO purifiers, roughly at a level with the gasoil purifier which is immediately aft of the gasoil oil settling tank. The high pressure lines above floor plate level are of relatively small diameter. It seems to be the consensus that a leak above floor plate level has a better chance of finding a source of ignition than a leak below floor plate level. Any conclusion as to the initial nature of the fire must be consistent with the evidence which I have already described as to the inspections in that area which were specifically directed to the ascertainment of leaks, and with the absence of any report of a really catastrophic leak causing oil to gush out which surely would have been visible to or audible by at least someone at some stage. There is no evidence of any event such as excessive vibration which might have given rise to a catastrophic failure, and no evidence to suggest that a catastrophic failure was likely to occur spontaneously. It would be unrealistic and naïve for me to proceed upon the assumption that the inspections by the Third Engineer and the Oiler were necessarily carried out with meticulous care. On the other hand I think I should credit them with sufficient diligence to have detected and to have taken note of anything untoward which was obvious. I therefore conclude that the fire is likely initially to have been relatively small, albeit producing a disproportionate amount of black smoke in a confined space, the result of a relatively small leak which may gradually have become larger but which is unlikely to have had the capacity to build up a large reservoir of oil on the floor plates or in the savealls before the fire began. This is I think of some importance in evaluating the likelihood of there having occurred the sequence of events put forward by Dr Foster. I regard it as a very real weakness in his evidence that he had given no or no adequate consideration to what were the realistic possibilities so far as concerned the origin of the fire and, in particular, the initial source of fuel. This was of obvious importance to an evaluation of the realistic possibility of the fire thereafter developing sufficiently rapidly to give rise to the various phenomena which Dr Foster posited. It is clear that Dr Foster saw it as his role to suggest ways in which the fire might plausibly have developed so as to lead to the loss of the vessel. There is nothing inherently objectionable in this approach so long as the expert keeps his feet firmly on the ground and keeps a careful eye on the question whether what he is proposing is likely to have occurred. I think that in the important respect which I have already described Dr Foster deprived himself, or was deprived, of the ability properly to keep this perspective in mind. Since he had given no thought or no focused thought to the initial source of fuel for the fire he was in no real position to form a view as to the likelihood of the build-up of fuel sufficient to fuel a fire which could have the effect of causing the bronze valves mounted on the aft face of the gasoil tanks to fail within such timescale as was required to provide a further source of fuel to account for such phenomena as were observed and were consistent with a rapid spread of the fire, e.g. bubbling of paint on the port side of the vessel.

[41]The Claimants' marine engineering expert was Mr Bevis. One of the most important areas covered by his evidence was to explain how fire might plausibly have given rise to an ingress of seawater, the most obvious contender being of course introduction through a breach or breaches in the integrity of the seawater circulating system. Mr Bevis told me that he did not, when preparing his first report, address the question whether it was likely that a fire would lead to the loss of the vessel within 15 hours. Mr Berry evidently thought that it was to his advantage to attempt to demonstrate that Mr Bevis must in fact have given thought to the likelihood of the fire causing the vessel to sink within the required timeframe. By showing to Mr Bevis what he had actually said in his first report which recited that the vessel had sunk 15 hours after the outbreak of fire and expressed the view that flooding was probably the result of damage sustained as a consequence of the fire and explosions Mr Berry persuaded Mr Bevis to retract his earlier answer, which he said must have been mistaken. He went on:- "You asked me whether my first report had addressed that and I gave you an incorrect answer, my Lord." It might be said that the very fact that Mr Bevis could be in any doubt about a matter which was, or should have been, so central to his work itself speaks volumes. Mr Berry wished to draw attention to a different point, the fact that Mr Bevis had not in his first report shared with the reader his view, which he said he had always had, that the flow rates which would be produced by the most obvious mode of failure, flange failure, would be nowhere near enough to sink the ship within 15 hours and yet he had not in his first report identified any other mode of failure likely to have been caused by fire or explosion. As to this latter point Mr Bevis said:- "I certainly had it in mind when I wrote my supplementary report, I did not identify it in my main report simply because the mechanism of leakage had not been addressed at that stage." Again I am afraid I have to record that the picture revealed was really very unsatisfactory. Notwithstanding the plain language of the first report to which Mr Berry drew attention, it is abundantly clear to me that Mr Bevis had not in fact when he wrote that first report given any or any adequate consideration to the question whether it was likely that the fire would cause the vessel to sink within 15 hours. He said in evidence that he believed that the fire could sink the vessel within 15 hours, but that he had not paused to consider whether that was in fact a likely outcome is underscored, in my view, by the fact that he discussed no mechanism which would actually give rise to a leakage of the required volume within the required timescale. The evidence which Mr Bevis gave left me in no doubt that he had not, when he wrote his first report, had in mind fracture of the seawater piping. It is plain to me that he in fact assumed that the fire had caused the ship to sink. It may have followed logically from his thought processes that some mode of failure other than flange failure must have been involved, but I do not believe that Mr Bevis at that stage in the exercise gave any consideration either to what it might have been, or more crucially, to whether it was likely to have occurred. This is perhaps not altogether surprising in view of the fact that seawater piping fractured, as opposed to leaking through flanges, in consequence of fire was completely outside the experience of Mr Bevis, as indeed it was of Dr Foster, who is vastly experienced in the investigation of the aftermath of fires. Indeed Dr Foster felt it inappropriate to propose forms of failure of the seawater system more catastrophic than failure of flanges. I shall have to discuss hereafter the evidence which Mr Bevis in fact gave about the possibility of fire induced fracture in seawater pipes, but it was to say the least confused. I do not believe that Mr Bevis had any very clear idea of what it was that he was proposing.

[42]Dr Foster was not an adherent of the pipe fracture theory so it was inevitable that he could not account for more than a fairly limited leakage through flanges which might have been in progress for something of the order of 30 minutes or so by the time of abandoning ship. Equally importantly however he could not account, and in all fairness he did not attempt to account, for water forward of the engine room bulkhead by the time the ship was abandoned. He put forward various theories whereby the watertight integrity of that bulkhead might have been breached, including a persistent oil fire up against it causing buckling or an explosion impelling towards it at high velocity a piece of pointed steel of sufficient size to cause significant damage. However the more prominent of these theories involve damage at high level. They would not lead to the introduction of water into No. 4 hold unless and until there was a very substantial head of water in the engine room. Even Mr Bevis thought that flooding of the engine room would be at only a very early stage by the time the ship was abandoned. An increase in midships draft of the order of 1m or even rather less by the time the ship was abandoned is consistent only with the introduction of seawater into some space forward of frame 36. There is simply no scientific basis upon which I could conclude that that is likely to have occurred in consequence of damage caused by fire or explosion, although again it is for consideration whether I am driven by my overall evaluation of the entirety of the case, including underwriters' acceptance that the fire was accidental, to the conclusion that it must in fact have occurred in that manner.

[43]I should mention that some of the criticism of the Defendants' fire expert Dr Bound was in my judgment also justified. Some of the criticism was not. He had prepared his report, as asked, on the assumption that the fire started accidentally. He included in his report the standard form recital that the assumptions that he had made were not unreasonable. I am not sure that he necessarily appreciated the significance of that recital in relation to this assumption, which was not his own assumption based upon any evaluation of any material but was an assumption which he had been asked to make. However it is obvious that the assumption carried with it certain necessary limitations as to the manner in which the fire might have started, if the evidence of recent inspections of the area was to be taken at face value and having regard to the presence of personnel in the engine room. Difficulties arose when the premise of questions put to Dr Bound involved a fire far more developed in its scale and intensity on discovery than Dr Bound plainly regarded as consistent with it having started accidentally. I believe that Dr Bound felt himself inhibited in answering questions of this sort by the convention on the basis of which the case was being conducted. However I do believe that Dr Bound was at times inclined to place disproportionate weight upon an isolated piece of evidence, without regard to the bigger picture, although here again all of the experts faced a dilemma. All of them were supplied with all of the evidence obtained from the crew, not all of which was consistent, as I have already pointed out. The expert cannot in such circumstances prefer one piece of evidence to another because that is the province of the court. Obviously an expert must in such circumstances make clear the factual assumption upon the basis of which any opinion is expressed. He runs the danger of criticism however if he apparently places more reliance upon pieces of evidence favouring his own side than on those tending in the other direction. As in all things a balance is required. Whilst understanding the nature of the difficulty I am not sure that Dr Bound always achieved quite the appropriate balance. Again I would not wish to be unduly critical. The experts were engaged upon an unusual exercise from a starting point not of their own making.

[44]The evidence of Dr Foster and Mr Bevis was directed to showing how, consistently with the observations of the witnesses, the fire might have developed in such a manner as to cause the vessel to sink, unaided by any other mechanism, within 15 hours. I have already commented upon some of the shortcomings in the exercise which they conducted. Furthermore in so far as Dr Foster's sequence of events was tailored to account for 3 distinct explosions before the vessel was abandoned, my findings of fact involve that any explosions prior to that which prompted final abandonment were of perhaps lesser significance than Dr Foster may have been inclined to attribute to them. That is not a matter for criticism of Dr Foster. He made clear the factual basis for his opinions. Nor do my findings of themselves preclude matters having developed as Dr Foster suggests that they might have done. Furthermore there was on any view by the time of abandonment a well developed fire which had spread from the confined area in which it was first observed right across the engine room so as to cause bubbling of paint on the outside of the shell plating on the port side and indeed heating of the deck plates. Even if I regard Dr Foster's sequence of events as implausible that does not remove the fact that a fire which is accepted to have started accidentally had in a short time found sufficient fuel to sustain and augment it. That said, the main significance of Dr Foster's sequence of events is not that he sought thereby to provide an explanation for what he thought were the observed manifestations of the fire but rather that he sought, subject to the limitations which I have already indicated, to identify a means whereby the fire and consequent explosions could cause a loss of the watertight integrity of the seawater system and the engine room forward bulkhead at frame 36.

[45]Dr Foster proposed a sequence of events as follows. An intense fire on the purifier flat after half an hour or so caused a failure of a bronze valve on the aft side of one of the gasoil service and settling tanks. The oil thus released migrated towards the fuel oil service and settling tanks, which are immediately to the starboard of the centre line, causing a similar failure of a bronze valve on the aft side of the settling tank. This self-same fire must not however have induced an explosion in the always explosive atmosphere above the fuel oil service tank since the ullage space in that tank would be sufficiently small to ensure an explosion of relatively low violence which might however be quite sufficient to render the service and settling tanks common. Dr Foster would prefer that the settling tank remained intact with however its ullage space being reduced by leakage through the failed bronze valve so as to permit the build up of a sufficiently large explosive atmosphere in the tank, consistent with the loud explosion associated with smoke at the aft end of No. 4 hold heard before the Master and Chief Engineer abandoned ship. This high violence explosion must cause preferential failure of the forward face of the tank, and in particular must project a sharp corner piece or some similar projectile at high velocity towards the engine room bulkhead, either piercing it or perhaps severing one or both of the air vent pipes which pass aft of the bulkhead from the double bottom fuel tanks to the main deck. One or other of these breaches thus allows smoke from the explosion to be visible on deck apparently at the after end of No. 4 hold. Meanwhile but of equal significance burning oil in substantial quantities coming from both the gasoil and the fuel oil daily tanks is falling into the lower part of the engine room so to create a fierce fire at low level which flourished notwithstanding the injection of CO2 so as to create the conditions in which there may be damage to the seawater pipes. I have already pointed out that Dr Foster did not himself feel it appropriate to suggest consequential catastrophic damage to seawater pipes but I must examine separately what Mr Bevis proposed in this regard.

[46]The experts were agreed that the initial fire on the purifier flat most likely involved fuel oil from the heater which would be heated to about 90 to 95 (C, above its flashpoint. The fire is likely to have been augmented at some stage by various inflammable materials in the vicinity, including 10 to 15 litres of cleaning agent and fuel or diesel oil mix in a sink on the purifier flat, between the IFO purifiers and the gasoil daily tanks, an open drum underneath that sink containing 20 litres or so of kerosene or diesel oil and three closed 20 litre drums containing fuel oil additive located aft of the fuel oil settling tank. Dr Bound accepted that the burning on the port side of the vessel which caused paint bubbling is likely to have involved more oil than can be accounted for by the initial leak – which must for reasons already discussed have been relatively modest – and the additional amount in the sink and in the drums to which I have just referred. There must therefore have been an additional source of fuel. I have already noted that continued operation of one or more of the generators is indicative that some or all of the quick acting valves did not close. It can therefore be assumed that the original leak may have continued.

[47]It is common ground that the bronze drain valves on the gasoil tanks could in time have failed in consequence of direct flame impingement, or of a combination of direct flame impingement and radiant heat flux. The relevant valves are in a cut-away section of the purifier flat plating, bounded by a 40mm upstand. A fire fuelled by oil collected on the raised platform plates would be unlikely to give rise to direct flame impingement – what is required for that is an accumulation of oil either in the saveall underneath the gasoil tanks or at the forward end of the saveall underneath the purifier flat. Accumulation of a significant amount of oil in either saveall would require either blocked or partially blocked scupper drains or arrival of the oil at a rate faster than that at which it could drain away, or a combination of the two. Although the scupper drains should not have been blocked, and although previous inspections should have revealed prior build up of oil in the savealls, I do not have much difficulty in envisaging that the scupper drains may have been blocked or partially blocked. Consideration whether it is likely that the necessary conditions obtained for the creation of(a) a fire underneath the valves, causing direct flame impingement on them for sufficiently long and

(b) a fire on the raised platform, sufficiently close to the valves and of sufficient intensity and duration to give rise to significant radiant heat flux is I believe simply speculation. It seems to be accepted that what would be needed would be a fierce fire for some 30 minutes or so (tens of minutes was Dr Foster's formulation) burning both in one or other of the savealls and not more than about half a metre distant on the platform. I find it very difficult to evaluate the realistic possibility of this occurring without knowing more about the nature of the initial leak. My conclusion that any leak consistent with an accidental fire must have been relatively small tends to militate against the likelihood of the bronze valves failing as soon as 30 minutes or so after the discovery of the fire. On the other hand the fire did continue to burn. In due course it spread in a manner which is consistent only with it having had a source of fuel other than simply the initial leak and the various drums of material stored in the vicinity. The gasoil tanks were in the immediate proximity of the initial location of the fire and the bronze valves on those tanks are, as Dr Bound accepted, the most likely to have been damaged in the course of the fire. I therefore conclude that given the length of time for which the fire is known to have burned and given the extent to which it is known to have spread it is likely that at some relatively early stage the bronze valves on one or other or both of the gasoil service and settling tanks failed, providing a further source of fuel.

[48]There was equally fierce debate on the question whether the fire is likely to have developed in such a manner as preferentially to have caused failure of the bronze drain valve on the IFO settling tank and yet not simultaneously to have caused a low violence explosion in either the IFO service or settling tank. Again the failure might be caused by direct flame impingement or a combination of direct flame impingement and radiant heat flux. If it were to be supposed that the failure was caused by radiant heat flux alone then the timescale involved would be comparatively and significantly extended. It was not suggested that failure through exposure to radiant heat flux alone was plausible within the required timescale. Again the inquiry is hampered by lack of any knowledge of the nature and size of the initial leak. An evaluation of whether a fire of the type observed starting in a manner consistent with it being accidental could, if left unfought, extend to the point at which there was direct flame impingement on the aft side of the IFO tanks is even more speculative than the similar inquiry in respect of the valves on the aft side of the gasoil tanks, which are at least very much closer to the original seat of the fire. What would be required for either direct flame impingement or radiation from a fierce fire half a metre away from the valve is a significant accumulation of oil against the upstand on the floor plates to the aft of the tanks. This is in principle unlikely to have occurred because there are scupper drains in the upper machinery flat designed to prevent it occurring. These scupper drains are additional to and not to be confused with the scupper drains internal to the purifier and oil tank savealls. Moreover the vessel was trimmed by the stern and this theory requires the oil to migrate forward. There is a further difficulty arising out of the location of the IFO settling tank bronze drain valve. In fact the precise arrangement is not entirely easy to deduce from photographs of the equivalent location on the sister ship – and of course the assumption must be made that the arrangements were there similar. Close examination of photographs taken from different angles revealed that the perspective from which a photograph was taken could give rise to a most misleading impression of the positions of two pieces of structure relative to each other. Subject to all those caveats the relevant valve appeared to be of all of the valves under consideration the least susceptible to either direct flame impingement or radiation. It is afforded at any rate some protection by two vertical pipes and one horizontal pipe and by a large pillar. It is difficult to imagine flame impingement on the valve as a result of fire on the floor plates. Dr Foster did not in his evidence offer any attempt to evaluate the contribution of radiant heat. Dr Bound did. Because of the many imponderables involved the exercise was inevitably somewhat unscientific and the outcome is in any event entirely dependent upon the intensity of the fire which one assumes to have been burning on the floor plates. Again all that I can reliably conclude is that the nature of the arrangements was such that the valve was to an appreciable extent shielded from the effects of radiant heat and that this renders it even less likely that this valve failed in consequence of the fire.

[49]However if the fire at the aft end of the fuel oil tanks was sufficiently intense to cause failure of the relatively protected bronze drain valve on the settling tank then it has to be considered whether that would not in all probability have lead to an early explosion of relatively low violence in either the service tank or the settling tank. It was common ground that the atmosphere in the ullage space of the service tank would be at all times explosive. In the light of that agreement it is unnecessary for me to attempt to decide whether it is likely that sufficient explosive vapour would be drawn from the service tank to the settling tank to render the ullage space in that tank explosive too. Dr Foster accepted that a fierce fire of the intensity required to cause bronze valves to fail would be likely to lead to the earlier cracking of the service tank sight glass which could very well lead to a low intensity explosion in the service tank. Quite what damage that would cause it is impossible to say. As Dr Foster pointed out, it is not inevitable that a small explosion in the service tank would breach the integrity of the settling tank, but it is an obvious possibility which Dr Foster recognised that it would cause such damage to the bulkhead between the two tanks as would preclude the persistence or development of an explosive atmosphere in the settling tank.

[50]The fact remains however that an explanation has to be found for the loud explosion which prompted final abandonment of the ship. Of course the magnitude of that explosion can only now be assessed as a result of the subjective evaluations of it to which I have already referred. However if this was a loud or significant explosion and if the smoke associated with it came from the No.4 access hatch or from the vent pipes on deck then Dr Bound for his part was not able to offer any explanation for it other than an explosion in the IFO settling tank. That in turn creates another difficulty as to the precise means by which smoke reached the point at which it was apparently observed. I understood Dr Foster to accept that if the access hatch and the weathertight door were closed it would require an explosion actually in the hold to disrupt them so as to permit smoke to appear. The only realistic mechanism for the appearance of the smoke spoken of by the witnesses, assuming their perception as to where it was coming from to be correct, is that the explosion caused a breach in one or more of the three vent pipes immediately aft of the bulkhead. These are constructed of 6mm steel, material similar to that of which the fuel oil tanks were likely to be composed. The explosion could have caused part of the forward bulkhead of the tank to go through one of these pipes, but that of course presupposes that the forward face of the tank or perhaps the roof failed in preference to one of the other sides. This again seems to me to enter the realms of speculation. It is obvious that the tank would fail at its weakest point. There is simply no way of knowing where any weld weaknesses might have been. Mr Bevis and Mr Todd were agreed that failure was likely to occur in way of tank weld seams. The aft face would be exposed to the direct effects of the heating, so one would expect this to expand most, placing stress on the welds attached to it. Against that the aft face had more fittings and attachments than the other faces, including the manhole, the overflow pipe, and the tank valves. They might have the tendency to add stiffness to the aft face, although Mr Todd pointed out that cutting out a piece and replacing it may simply restore the original strength – or it may introduce still more weak points. It is quite impossible to resolve this debate in any principled manner. All one can say is that an explosion in the ullage space of the settling tank could have caused steel work from the tank to be displaced forwards.

[51]The next question is whether this displaced steel work could have caused either a breach in the bulkhead at frame 36 or a breach in the vent pipes or both. Again it is not possible to say any more than that this could have happened. No-one could possibly sensibly assert that it was likely and no-one did. Dr Bound thought that both were low probability outcomes. So far as concerns piercing the bulkhead, Dr Foster seemed only to envisage that this might be caused by a sharp corner flap or something similar, and there would have to be the fortuity of the flap not being arrested by any pipes or stiffeners and striking the bulkhead at the required angle. So far as the pipes were concerned I was rather more persuaded by the notion that they might be crushed or distorted – simply bent, rather than severed, but severance cannot be ruled out. Nor can a breach in the bulkhead but I conclude that any breach caused in this manner is likely to have been relatively small and it would of course be at high level. The significance of this is that it would not afford a means whereby water could enter hold No. 4 until the engine room was almost totally flooded and even then it would be unlikely to permit any very rapid flow of water from the one compartment to the other. Of course I cannot rule out that a sustained fire burning against the forward bulkhead may have caused significant distortion to it, which could have caused a loss of watertight integrity but Dr Foster did not seem to me to place this in the forefront of what he regarded as likely to have occurred.

[52]This brings me to what is in my judgment the single greatest obstacle to acceptance that it is likely that the fire caused the vessel to sink in 15 hours. This is the question how sufficient water can have entered the vessel within the required timescale in consequence of damage caused simply by fire or explosion. I have already alluded to the difficulties which this part of the case posed for the Claimants. In fact the problem is two-fold, adding a further dimension to the sequence of events which must occur at the upper level of the engine room in order to render possible let alone probable a sinking solely by the effects of fire and explosion. In that regard I should mention that it was not contended that any explosions which may have occurred might plausibly have caused damage to seawater pipes. What is relied on is the effect of fire. However the relevant seawater pipes were all at low level in the engine room and so it is first necessary to posit a fierce fire at that low level fuelled by oil released from the upper part of the engine room. It was common ground that such gasoil as may have escaped through failed bronze drain valves would be insufficient in itself both to fuel the fire of the necessary intensity required to lead to a similar failure at the IFO tanks and to fuel a fierce fire at lower level of the intensity required to cause significant damage to seawater pipes. So some sort of failure at an IFO tank, whether of a drain valve or an explosion, is required to account for the presence of a sufficient quantity of oil at the lower level. There was a great deal of discussion about the quenching effect of the CO2 at this low level, and in particular the disposition of the CO2 nozzles and the fact that CO2 is heavier than air. Dr Bound thought that it was likely that the CO2 would be effective to extinguish any early fire at low level, Dr Foster accepted only that it was possible, although I thought that he was inclined to downplay the effectiveness of the CO2 at this lower level. However he did point out that if there is a large fire somewhere in the engine room that will have the effect of stirring up the atmosphere and more of the CO2 will be drawn away with the hot gases. The only realistic conclusion that I can draw from this is that the relative effectiveness of the CO2 at low level means that it would be surprising if a fierce fire were permitted to persist at that level. However what is much more difficult to assess is the effectiveness of the earlier released CO2 to deal with later released large quantities of oil already burning when it reached the lower levels of the engine room, assuming this to have occurred. There are few plausible sources of ignition at the lower level with the main engine having been shut down at an early stage and the diesel generators not in the likely line of any oil cascading down from above, but I was persuaded that if there were a substantial release of fuel oil, perhaps consequent upon an explosion in one of the fuel oil tanks, then it is possible that some oil, perhaps in quite substantial quantities, could be burning as it went down through the scupper pipes or over the upstands around the upper platforms and that the CO2 would not by that stage in the sequence be effective to quench the fire. The witness evidence is broadly suggestive of the CO2 being initially effective in subduing the fire but of that effect being relatively short-lived. However I think it also follows that any burning oil introduced to the lower levels in this way would have had very little opportunity to cause substantial damage by the time the ship was abandoned. This is another difficulty in the Claimants' way.

[53]I turn to deal with the mechanism by which sufficient damage could be done to the seawater pipes to give rise to full bore leakage or something approaching it. Nowhere in his written reports did Mr Bevis explain how this could come about and I am afraid that I derived the clear impression that he was describing a phenomenon of which he not simply had no experience, as he accepted, but also no understanding. In his first report Mr Bevis said this:- "Assuming there was no deliberate flooding of the engine room, then the most probable source of ingress on this occasion was fire damaged seawater pipes that leaked or ruptured allowing the engine room to flood. I am satisfied that such a scenario can occur." In the Joint Experts' Report the following was said:-"14. It is agreed that an intense and prolonged fire can result in fractures in seawater piping. On the assumption that there were no pre-existing defects, ART [Mr Todd] is of the opinion that such fractures are unlikely to result in full bore pipe leakages or any significant proportion of full bore flow. CJB [Mr Bevis] is of the opinion that such fractures could give rise to significant ingress of water.

15. We agree that, when subjected to an intense and prolonged fire, dry or air locked seawater pipes are likely to fail more quickly than pipes filled with, or circulated by, water." There was some discussion of the distinction between full bore separation and full bore leakage. Obviously the former may not lead to the latter and possibly the latter is never strictly achieved since there will always be some obstruction and there is the so-called "venturi" effect to take into account when calculating the rate of flow through an aperture. I regarded this discussion as a distraction. I do not read the joint report as containing any support for the notion that the pipes would sustain damage amounting to or approaching full bore separation. Mr Bevis seems to have chosen his language carefully to avoid supporting such a case, confining himself in contradistinction to Mr Todd's reference to full bore pipe leakage or significant proportion of full bore flow to a reference to "significant ingress of water."

[54]In his supplementary Report Mr Bevis was a little more forthcoming as to the mechanism. In paragraph 14.3 he said:- "Sections of pipe and valves that were dry when exposed to fire, for example scupper pipes, and sections of sea lines incorporating loops or closed ends where vapour may be trapped as the pipes are heated, are likely to have sustained damage." At Paragraph 15.6 he said this:- "The Chief Engineer has confirmed that there were no temporary patches on the system and sections of pipe had been renewed at the last maintenance period in Piraeus. Nevertheless, such a system when exposed to a severe fire may sustain damage as a result of stresses imposed due to expansion or fracture caused by shock. I would expect pipe failures and significant leaks." Thus the mechanism proposed was "stresses imposed due to expansion" or "fracture caused by shock." For what it is worth I do not read paragraph 18.4 of the same Report as supporting the notion that these mechanisms would give rise to full bore separation. The first two sentences of that paragraph read:- "Another type of failure would be fracture of a seawater line or valve. Neither is likely to result in full bore leakage but significant quantities of water would be involved." 55. The point was then developed in oral evidence in a manner which I found most confusing. On day 12 of the trial there first took place the following sequence of questions and answers: - Q. How on earth does a fire cause full bore failure in a seawater pipe? A. What happens, I have not seen a fire damaged pipe, seawater pipe as such, but I have seen a lot of seawater pipes which fail as a result of corrosion, vibration, shock, and the process is very similar. Q. That's a completely different mechanism? A. No it's not. What happened here is in those cases the pipe detaches from the weld, usually the weld to the flange at the end of the pipe connecting it to another fitting or another pipe, the weld being preferentially attacked by corrosion, wasted, weakened, and the pipe is then either pulled back from the joint or displaced vertically or actually from the other section of pipe. If the pipe drops you will get a partial full bore failure, and that's why I have tried to demonstrate in the table from coefficients going down from 5,4,3,2,1, and then 0.05. Q. You are talking about a case of extreme corrosion where the pipe detaches from the weld. That's nothing to do with fire? A. It is, that's exactly what I expect to happen in a fire. What is going to happen in a fire is that once the seawater in the system stops circulating, the water contained within the pipe will be exposed to severe heat until that water evaporates. The temperature of the pipe wall, because the pipe is only thin, will not rise above 100ºC theoretically anyway. As soon as the water turns to steam, the heat transfer which you have hitherto relied on to conduct heat away from the pipe stops, the temperature of the pipe can increase. As the temperature of the pipes increase, differential expansion of the hot sections and the cold sections of the pipe will cause stresses to be placed on the pipe work which can and probably will separate those pipes from welds." This seemed to involve differential expansion causing separation of welds, or of pipes from welds, but later on the same day Mr Bevis referred to shock:- "A. I haven't changed my evidence, I think this scenario is unlikely to occur unless the pipe was struck by a foreign object and completely sheered off, but in the scenario I have given you, which is heated pipe, expanded, exposed to shock, I think this sort of scenario is highly likely to occur, indeed probable to occur."

[56]The matter was further explored on the following day of the trial, day 13. At first, the mechanism seemed merely to be differential expansion. Thus Mr Bevis said this:- "Q. Now the second mechanism that you have mentioned of fracture, the way that works, if it works at all, is where you have got your loop or stagnant end, you boil the water away and how that might cause a fracture then it is a differential expansion of the steel, the dry bit gets hotter, the cool bit is cooler and the steel moves relative to each other? A. Expansion of steel, steel has a coefficient of expansion so the higher the temperature you expose that pipe to or that steel to, the longer it gets and if the pipe is constrained, and these pipes are clipped in place to stop them moving about, then significant loads can be put on the pipe-work and I envisage that this will cause fractures. Q. But the mechanism is differential expansion, one piece of steel that's getting hotter than another piece of steel, putting stress? A. Let's say one piece of steel getting hotter causing expansion that can't be accommodated elsewhere in the system either because the other part of the system is cold or because the pipe-work is retrained or constrained in some way." However later on the same day shock was reintroduced, when Mr Bevis said this:- "A. No the pipe-work is clipped, it is held by brackets, restrained, and a section of pipe that is heated and allowed to heat until it is dry will grow in length and that expansion can't be accommodated by adjacent sections because they are restrained, so sections of pipe, pipe connections, pipe flange connections will be subjected to stress. Q. So it is not part of your thesis that differential expansion of the hot section and the cold section will…. A. There will be hot and cold sections and, as the water gets into a hot section I can foresee the steam collapsing condensing very quickly, causing shock in the pipe system. I can see that happening during this process of heating, but essentially what I'm telling you is that I believe the stresses will be imposed, we are talking, essentially working towards explaining how fractures occur, and I believe that the predominant factor will be heating of pipes which are not able to accommodate the expansion of the section that is heated." He clarified however that he saw as the most obvious point of failure the point at which the pipe is welded to a flange. I asked him the question "it pulls away because the restrained system is unable to accommodate the expansion?" to which Mr Bevis replied that that was correct. However it appeared that the mechanism was not quite so straightforward. Thus a little later Mr Bevis said this:- "A. I have seen this scenario where pipes, seawater pipes have failed. I believe in my supplementary report I address the section on corrosion of seawater pipes to show that they would not, that sections of them would be weakened by corrosion, especially on a vessel of this age, and in the past I have seen seawater pipes have failed as a result of shock or vibration where they have detached from a weld, and I consider the failure, the effect of the failure will be very similar. Q. You are constructing an analogy with a corroded pipe that has had shock? A. That's correct. Q. And saying fire can do the same thing that? A. The effect of fire can do the same thing. Q. By differential expansion? A. By fracturing the pipe as a result of other thermal stressing." Mr Bevis went on to summarise it as expansion which could not be accommodated in the system. It transpires that this theory also required there to be relatively elderly portions of the seawater piping system since new steel would have sufficient ductility to accommodate expansion by bending without breaking. This vessel had had large scale renewals of seawater piping in 1996 although no doubt that did not involve the entire system. Again I asked Mr Bevis to clarify:- "MR JUSTICE TOMLINSON: Can I be entirely clear about this because we are using the word fracture somewhat loosely and, indeed, your joint agreement with Mr Todd talks about fractures in seawater piping. I perhaps wrongly read that as meaning fractures in the material of which the pipe is made but that's not what you are talking about at all, is it? It is not a fracture of the pipe, it is a parting of the weld? A. I think that's what I am talking about, my Lord. The weakest part, no, that's wrong. The weaker parts generally tend to be the welds, particularly at flange joints, the reason for that is that the weld metal, the weld may not have been completed as perfectly as one would like to anticipate that they always are. But in addition to that, in its assembled state as the water flows through that boiled flange it is extremely rare that the gasket, the weld, and the inner bore of the pipe would be a smooth transition and there will be turbulence generated there and that turbulence leads to aggravated corrosion." Mr Todd did not accept that there would be present the aggravated corrosion which Mr Bevis seemed to be positing as part of the mechanism.

[57]On any view the mechanism proposed, of which neither Mr Bevis nor Dr Foster claimed ever to have seen any example, involved boiling off of water into steam and a long period of dryness to cause the expansion which could not be accommodated. Mr Todd did not see the potential for this to occur. The system, which is of 300 and 250 mm pipe-work, does not have sharp bends in it. The system was open at both ends, inlet and discharge. The generation of steam would generate flow within the system and any water boiled off into steam would constantly be replaced by incoming water. Mr Todd did recognise the possibility for fractures of piping in two situations. Heating of the pipe-work could give rise to physical distortion which could possibly give rise to small fractures, but once the fractures occurred the stress would be relieved, so that full fracture running the whole way round the circumference of the pipe would be unlikely. He in fact saw as a more likely source of fracture dry pipes above the waterline being heated by intense fire and being immersed in rising seawater, but obviously such fractures could have played no part in the introduction of seawater sufficient to flood the engine room. Mr Todd's evidence on this point was consistent from first to last and his oral evidence was clear and convincing and not in any way shaken by cross-examination. I thought that Mr Todd gave his evidence carefully and thoughtfully, giving me the benefit of his long experience and keeping his feet firmly on the ground. Mr Todd thought it was highly improbable that there would be a fracture or rupture or detachment of a pipe sufficient to give rise to full bore leakage, or anything approaching it. He accepted that small cracks could occur but did not think that complete fracture was likely in pipe-work which was of normal serviceable condition and not on the verge of failure. I prefer Mr Todd's evidence on this aspect to that of Mr Bevis. The evidence of Mr Bevis on this point was most unclear, deriving from the circumstance that he had not originally considered how sufficient water could be introduced in sufficient time. Indeed when asked why he had not dealt with the suggested mechanism in his first report Mr Bevis replied:- "In my first report I was asked to address certain issues, this was not one of them." In order to make good the deficiency Mr Bevis came up with an ill-thought through theory which went beyond his and anyone else's experience. I cannot understand on what basis Mr Bevis felt able to say, in the experts' joint memorandum:

"CJB is of the opinion that 15 hours between the outbreak of an engine room fire and sinking is not exceptional."

The statistical evidence produced by Mr Bevis himself militated against this conclusion as did both his own and Dr Foster's lack of any experience of a fire induced catastrophic failure of sea water piping.

[58]Even if I had not concluded that it is unlikely that the fire alone could have been responsible for the introduction of sufficient water to sink the ship within 15 hours, still on the basis of the evidence of Dr Foster and Mr Bevis I would be unable to conclude that it is likely that the fire alone was responsible for ingress of such sea water as must have entered the vessel by the time the ship was abandoned. In that regard even Mr Bevis accepted that he could not suggest that the flooding was at more than an early stage at the time of abandonment. Dr Foster did not suggest that any fierce fire would by then have been raging for more than about an hour at most. Even on his best case Mr Bevis did not think that there would have been ingress of water of more than about 300 tonnes by the time the ship was abandoned. 300 tonnes of water in the engine room would have made no appreciable or noticeable, certainly not a "practical," difference in the midships draft. It would be consistent with a midships draft of about 9.6m.

[59]Furthermore wholesale acceptance of the evidence of Dr Foster and Mr Bevis could not lead to a conclusion that it is likely that, in consequence of the fire, any significant quantity of water could have found its way into No. 4 hold by the time of abandonment. A high level breach of the bulkhead would not lead to leakage of water into No. 4 hold until the engine room was nearly completely flooded. A breach of the vent pipes would similarly be at high level, and would incidentally lead to the eventual introduction of water into the double bottom fuel tanks, not No. 4 hold. There is a sounding pipe leading to No. 4 hold bilge which might have been vulnerable to damage from an explosion in one or other of the gasoil daily tanks. That again would be high level damage, and any water leaking down this small pipe would encounter the resistance of the rock phosphate cargo stowed over the bilge well in No. 4 hold. At a late stage Dr Foster raised the possibility to which I have already referred of the bulkhead at frame 36 fracturing in consequence of heavy buckling or distortion caused by an intense fire raging against it, perhaps at a low level. This struck me as a pretty desperate throw, and to his credit Dr Foster pointed out that he had seen many bulkheads that had been quite heavily buckled by fire and yet had not fractured. Indeed Dr Foster also accepted that explosions in fuel tanks could not be regarded as a usual consequence of a fire in an engine room.

[60]Thus the Claimants in my view have a two fold difficulty in explaining how there could have been any significant quantity of water forward of frame 36 by the time of abandonment. Yet consistent with my finding that there was a practical increase in draft there must at that stage have been water forward of frame 36. I have concluded on the basis of the evidence that the increase was probably of the order of about a metre. However it makes no significant difference for present purposes if the increase was in fact only of the order of about half a metre or even less than that. A midships draft of about 10m would require about 1600 tonnes of water in the engine room if that were the only flooded space. However that quantity of water in the engine room alone would put the Port of Registry markings nearly half a metre below water. I place no reliance whatever upon the suggestion that the height of the Port of Registry markings above a moving water line could be accurately estimated from a distance of about 500m to be about 30cm. But the reported observation at least involves that the Port of Registry letters were clearly visible and distinctly above the water line. That observation when combined with the observation that there had been a practical increase in midships draft is consistent only, as I understood the agreed evidence of the naval architects, with there having been water forward of frame 36 on abandonment. How much of course depends upon what was the midships draft and what was the height of the Port of Registry letters above the waterline. If the midships draft was only 10m and the Port of Registry letters were only 30cm above the waterline, the water required forward of frame 36 would be 700 tonnes with a further 770 in the engine room. A midships draft of 9.75m I would not regard as consistent with the observation of a practical increase. But even on that assumption the Port of Registry letters would need only to be more than 45cm above the waterline for there to be a requirement for water forward of frame 36. None of this can be precise. I am however satisfied that the reported condition of the vessel on abandonment is consistent only with water forward of frame 36 both the presence and the required quantity of which cannot be regarded as the likely consequence of the fire and explosions alone.

[61]In the light of these conclusions I do not believe that I need to consider further the possibility of eventual downflooding into no. 4 hold through the weather tight but not watertight openings. I have already accepted that there is a risk of this occurring when the relevant opening is less than 30cm above the static equilibrium waterline. However even then the process would be slow, although I accept that it would accelerate as the freeboard decreased. At the early stages of downflooding any water ingress would be dependent on waves washing over the deck and the initial rate of ingress would be small, assuming the watertight hatch to be in fact closed and in reasonable condition. It might ultimately be overcome and permit larger leakage but this would be a gradual process. As I understood it, and on the assumption that there was no water forward of frame 36, the risk of downflooding at no. 4 would not arise until the engine room and all spaces aft of frame 36, including those aft of frame 10, were completely flooded. Flooding of the spaces aft of frame 10 is disproportionately effective in reducing the aft draft for the obvious reason that it is the space at the aftermost end of the vessel. But the most plausible mechanism of leakage into this space is leakage past the steerage gear gland. As to this the naval architects said in their joint memorandum:- "It is agreed that leakage past the steerage gear gland could possibly occur, but if this occurred it is likely to be a restricted flow rather than a significant ingress. The amount could not be reliably assessed owing to the assumptions that have to be made. In any event, leakage will occur only at a later stage with over 11m aft draft."

[62]In the light of my findings as to the condition of the vessel on abandonment I consider it unlikely that the ship sank in consequence of downflooding into No. 4 hold. The evidence does not in any event lead me to the conclusion that the effects of fire alone would give rise to sufficient water ingress to render likely loss of the vessel through downflooding into No. 4 hold within 15 hours of the outbreak of the fire.

[63]Notwithstanding the findings to which I believe the evidence, factual and expert, inexorably lead, I have none the less stood back from the case to ask myself whether I should conclude that the fire simply must be regarded as the likely cause of the loss. In so doing I have tried to have regard to the guidance to be derived from the appellate decisions, particularly The Popi M

[1985] 1 WLR 948

, The Marel

[1994] 1 Lloyd's Law Reports 624

, The Ikarian Reefer

[1995] 1 Lloyd's Law Reports 455

and The Kapitan Sakharov

[2000] 2 Lloyd's Law Reports 255

. In evaluating the expert evidence I have naturally borne in mind the following observation of the Court of Appeal in The Ikarian Refer at page 484:- "Where the owner's explanation requires a series of steps to happen in sequence, each of which is improbable or highly improbable, the explanation may become incredible, especially if some or all of the steps have to take place within a tight time-scale and involve one or two remarkable coincidences. As Mr

Justice Greer pointed out in The Ionna [1922] 12 Lloyd's List Reports 54:

"One improbability would not be sufficient to justify one coming to the conclusion that the event did not happen. But when there are two improbabilities the likelihood of it happening is still more remote, and when there are three it is more remote still."

However it is always possible that the remarkable coincidences did occur, that the unexpected did happen, and of course perfectly possible that matters developed in so unexpected a manner that the sequence of events from the initial outbreak of fire took a form which has simply not been contemplated by the experts. I have tried to form a view as to what did happen. I have been defeated in that task, and it is possible that it was always going to be hopeless once the trial took the shape it did. As I have already pointed out, no possible explanations for the loss alternative to that proffered by the Claimants have been explored in the evidence. I am certainly not in a position in which all the possible explanations except an improbable one have been eliminated. On the other hand there is here present the feature, absent I believe in all previously decided cases, that the underwriters accept that there was an accidental fire, which incidentally caused the CTL of the vessel. I suspect that, both as a matter of principle and in the light of the approach of Dillon LJ in The Marel at page 633, I should put wholly out of mind all possible suspicion of scuttling. Even so, I must in all fairness to the Claimants observe (a) that an accidental fire occurring after a scuttling was already underway would be a very remarkable coincidence and (b) that the evidence which I heard did not inform me of any plausible mechanism whereby the vessel could have been scuttled assuming that that enterprise started only after the discovery of the fire.

[64]I am left at the end of the day with a simple question. Does the underwriters' acceptance that the fire was accidental compel me to the conclusion that the accidental fire simply must have caused the vessel to sink on the basis that the evidence has thrown up no realistic alternative. Fires can obviously develop in wholly unexpected ways. Steel ships can react to external forces in unexpected ways, particularly when they are old. No-one has set out to demonstrate to me any plausible mechanism whereby water might have been introduced into this vessel and caused it to sink in a manner which does not involve that the fire, including of course any consequent explosions, was the proximate cause of the sinking. That notwithstanding, I have ultimately concluded that underwriters' acceptance that the fire was accidental does not of itself or when combined with the evidence compel me to the conclusion that the fire was probably the cause of the sinking. The fact is that I have concluded, on the basis of my evaluation of all of the evidence, that there must have been present on abandonment water, including forward of frame 36, the quantity and location of which simply cannot be regarded as the likely consequence of a fire which had started so soon before. On that basis alone I cannot reach a principled conclusion that the fire was probably the proximate cause of the sinking. In any event I cannot conclude that the fire was the probable cause of an ingress of water sufficient to sink the vessel within 15 hours.

[65]There remains the question whether I should regard the fire as one of two equally effective concurrent causes. Since the unexplained ingress of seawater is not shown to have been due to an excepted peril, the Claimants could recover on this basis if the fire was a proximate cause of the loss – see Arnould, Marine Insurance, 16 th Edition, at paragraph 775. Mr Hofmeyr proffered as a working definition of proximate cause the event, whether peril or exception, which in all the circumstances prevailing at the time of loss lead inevitably to the kind of loss in question – see Clarke on Insurance Contracts paragraph 25.4. On my findings the fire does not qualify as a proximate cause of this loss which it must be remembered is not simply a sinking but a sinking within 15 hours of the outbreak of fire. On my findings the fire alone would not have lead to the loss of the vessel within that timescale. It is in my judgment insufficient for the Claimants to show that the aggregate of the ingress of seawater caused(i) by fire and

(ii) by unexplained means caused the sinking. Nor is it in my judgment enough for the Claimants to show that the unexplained flooding was itself insufficient to cause the sinking. On the footing that the fire probably gave rise to some ingress of water, that must inevitably in one sense be so. But the Claimants have not in any real sense shown that the unexplained cause could not have accounted for sufficient water to sink the ship within 15 hours – the unexplained cause may have had the capacity to admit more water had not some water from another source, the fire, already occupied any relevant flooded space. However a finding of a contribution, without which the vessel would not have sunk when she did, does not lead to the conclusion that the contribution should itself be regarded as a proximate cause of the loss. On my findings this vessel would not have sunk when she did without a large unexplained ingress of water. On that basis the water contributed by the fire cannot itself be regarded as a proximate cause of the sinking. That I believe is sufficient to defeat the Claimants' claim on this basis. However I would if necessary go further. The water contributed by the fire may have been in one sense a concurrent cause. But I cannot conclude that it was of the same efficacy as the unexplained ingress. Where there are two concurrent causes, one insured and one not, although not excepted, I do not believe that the insured can recover unless the causes are of equal efficacy. Perhaps that is simply another way of saying that the causes must both be proximate. Standing back, I agree with Mr Berry that it would be contrary to common sense to find that whilst the fire could not cause enough water to enter to sink the ship, at any rate within the required time, nonetheless the insured can discharge the burden of proof by proving that the fire must have contributed some water. On that basis an insured contribution of 10 tonnes where 5000 were needed in all would suffice to enable the insured to recover. That is not a sensible result.

[66]Conclusion It follows that had it been necessary to consider the Claimants' claim for an actual total loss I would have concluded that that claim fails. However for the reasons which I have set out earlier in this judgment the Claimants are in my view entitled to recover from the Defendant underwriters in respect of the constructive total loss of the vessel. There must be judgment accordingly.

Cited in 1 later judgment

Kastor Navigation Co Ltd & Anor v AXA Global Risks (UK) Ltd & Ors 2004