Court of Appeal · Decided 19 March 2026

The insured did not prove the loss was caused by perils of the seas

Argoglobal Underwriting Asia Pacific Pte Ltd and others v Oversea-Chinese Banking Corp Ltd

Feedback & support

[2026] SGCA 14Case no. Court of Appeal / Civil Appeal No 18 of 2025

The decision, in one line

Not legal advice

The insured did not prove the loss was caused by perils of the seas

Summary by laws.sgParagraphs [5], [115]

There is no fuller plain-English summary of this judgment yet.

Ask about this judgment

Laws this judgment relies on

Grounds of decision

Official text from eLitigation

Introduction
[1] Not infrequently, ships in the course of their voyages are lost at sea. Such incidents call for investigations to determine the cause especially in the context of a claim under a marine insurance policy. How can a shipowner be expected to prove the cause of the loss where the vessel has sunk? Is there room to invoke Sherlock Holmes’s celebrated investigation theory that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth”? While there might be a certain attraction in its logic, the House of Lords in Rhesa Shipping Co SA v Herbert David Edmunds [1985] 1 WLR 948 (“The Popi M”) observed that it “does not accord with common sense” to find that an event is more likely to have occurred even if it is “extremely improbable” (at 956C). It remains the task of the court to undertake its fact-finding process in order to determine whether the burden of proof has been discharged.
[2] Instead, the law has developed a rebuttable presumption that the loss of a vessel was caused by perils of the sea if the shipowner is able to prove that the vessel was seaworthy and that the loss occurred in “wholly unexplained” circumstances. This would typically arise where a vessel has sunk, and the law recognises the enormous difficulty for the shipowner to establish the cause of the sinking, hence the “wholly unexplained” condition needs to be fulfilled.
[3] It is important to bear in mind that the rebuttable presumption is not intended to operate as an evidential tool of convenience. Therefore, where the casualty is capable of being explained and the shipowner elects not to take steps to investigate the cause, it would not be entitled to invoke the presumption simply because it would have failed to satisfy the “wholly unexplained” condition.
[4] This crucial distinction was brought to fore on the facts of this appeal. In this case, the vessel did not sink. It capsized and remained afloat for a number of weeks. During that period, the vessel was inspected but no step was taken to investigate the cause of the capsize and no evidence was adduced at the trial that the cause could not have been ascertained. The vessel was eventually scuttled by the shipowner after it had served a notice of abandonment on the insurers based on its claim that the vessel was a constructive total loss (“CTL”) and no offer for its scrap value was reportedly forthcoming.
[5] There are two pivotal issues in this appeal. First, has the insured proved that the loss was caused by perils of the sea? A related question is whether there is any room for the presumption to apply in such circumstances. Second, proof of an insured peril would only mean that the marine insurance policy is engaged. The insured would still need to prove that the vessel was a CTL. In the proceedings below, the High Court judge (“Judge”) found in favour of the insured on both issues even though no witness was called to prove the CTL. After having carefully considered the parties’ submissions, we disagree with the Judge on both findings though a reversal of either of them would have sufficed to allow the appeal.
Background facts
The Tow Voyage
[6] Oversea-Chinese Banking Corp Ltd (“OCBC”) was the mortgagee of the vessel, “TERAS LYZA” (“Vessel”). The Vessel was a jackup rig built sometime before 2018. Under a hull and machinery marine insurance policy (“MI”) issued by the five appellant insurance companies (“Insurers”) in respect of the Vessel, OCBC was the co-assured alongside the vessel owner Teras Lyza Pte Ltd (“TLPL”) and the vessel manager Teras Offshore Pte Ltd (“TOPL”). TLPL and TOPL will collectively be referred to as the “Teras Entities”. Their parent company was Ezion Holdings Ltd (“Ezion”). For convenience, where a distinction does not need to be drawn between the Teras Entities and Ezion, we will refer to them collectively as the “Owners”.
[7] The MI contained a “Cover Note” issued by marine insurance broker LCH Lockton Pte Ltd (“LCH”) to TLPL on 13 June 2017. Under Section [A], the hull and machinery of the Vessel was insured up to an insured value of US$56m and, under Section [B], for an “increased value and/or excess liabilities up to a value of US$14m”. The Judge found that Section [B] was void as a gaming or wagering contract. This finding is not challenged in this appeal and hence, only Section [A] is relevant.
[8] LCH issued an “Addendum No 1” to the Cover Note on 13 June 2017 under which each Teras Entity assigned their rights, title and interest in the MI to OCBC, which was named as the “sole loss payee”. Addendum No 1 incorporated the relevant notice of assignment and loss payable clause into the MI. On 29 August 2017, a renewal certificate was issued to Ezion confirming the insurance of the Vessel for the period of 1 August 2017 to 31 July 2018 (“Renewal Certificate Policy”).
[9] The Vessel was subsequently prepared to embark on a tow voyage from Vung Tau, Vietnam to Taichung, Taiwan (“Tow Voyage”) for delivery to a set of charterers. This was to be its maiden voyage. The Insurers agreed to insure the Vessel for the Tow Voyage, and this agreement was formalised in an “Addendum No 3” issued to the Cover Note on 6 June 2018 and an Endorsement No. 18183166/7 issued to the Renewal Certificate Policy on 20 June 2018.
[10] The Teras Entities undertook an internal feasibility study to assess the feasibility of a wet tow of the Vessel. Conducted by then-Deputy Director of Fleet Operations of TOPL, Captain Bjarke Norby Pedersen (“Cpt Pedersen”), the study concluded that a wet tow was feasible, subject to the flag state and other approvals as required. A surveyor from the American Bureau of Shipping (“ABS”) attended the Vessel on 25 May 2018. Thereafter, ABS issued TOPL a Certificate of Fitness to Proceed under Tow, a Short-Term International Load Line Certificate, a Class Survey Report, and a Statutory Survey Report.
[11] The Teras Entities first contacted and intended to engage Braemar Technical Services (Offshore) Pte Ltd (“Braemar”), a competent marine warranty surveyor (“MWS”) for the wet tow. This engagement was pre-approved by the Insurers. However, for reasons disputed by the parties, Techwise Offshore Consultancy Pte Ltd (“Techwise”) was eventually appointed as the MWS instead. Between 26 and 31 May 2018, Techwise issued a Certificate of Fitness for Towage, a Certificate of Approval for the Tow Voyage, a Suitability Survey Report and a Sailaway Attendance Report.
[12] On 30 May 2018, the Vessel, under the tow of the Teras Eden (“Tug”), sailed away from Vung Tau, Vietnam. On 5 June 2018 at 3.45pm, the Vessel developed a list to port and trimming by the stern and at 5.50pm the same day, the Vessel capsized.
Events after the Vessel’s capsize
[13] Salvors from Resolve Salvage & Fire (Asia) Pte Ltd (“Salvors”) were contracted to salvage the Vessel on 15 June 2018 by invoking the Special Compensation, Protection and Indemnity Club (“SCOPIC”) clause of Lloyd’s Standard Form of Salvage Agreement. The Vessel was towed to Batangas Bay, Philippines while in a capsized state.
[15] The Insurers responded on 11 July 2018, disputing the claimed figures for being too high and stating that it was premature to conclude that the Vessel was a CTL (“11 July Email”).
[16] By an email dated 22 July 2018, the Owners recorded an impasse between the Protection and Indemnity Club (“P&I Club”), Owners and the Insurers over a number of issues (“22 July Email”). The Owners tabled their strong views, supported by the P&I Club, that the casualty was a CTL, although there was a general consensus that a proper diving inspection should be done to inspect the structural condition of the Vessel. The Insurers maintained their view that the Vessel was not a CTL.
[17] Between 23 and 25 July 2018, an underwater inspection of the Vessel was carried out by Solis Marine Consultants (“Solis”). Pursuant to the underwater inspection, Special Casualty Representative Situation Reports (“SCRS Reports”) were produced and sent to the Owners and the Insurers. The SCRS Reports set out various observations on the extent of damage suffered by the Vessel.
[19] The Salvors issued a report (“Dive Inspection Report”) on 4 August 2018, following several inspections between 23 July and 27 July 2018. The Dive Inspection Report detailed the damage suffered by the Vessel.
[20] On or around 4 August 2018, the Owners started making enquiries and received responses from various shipping brokers (“Ship Brokers’ Replies”) with respect to the scrap sale of the Vessel. However, there were no potential buyers for the Vessel as scrap. On 18 August 2018, the Owners informed the Insurers that no scrap buyers were interested and that the Vessel could only be disposed of (“18 August Letter”). The Vessel was then disposed of and submerged by the Salvors in deep water off the Philippines on 20 August 2018.
[21] Subsequently, OCBC and the Teras Entities mounted a claim under the MI. The MI incorporates the Institute Time Clauses (Hulls) 1.10.83 CL 280 (“ITC”), which provides that the insurance is subject to English law and practice. Clause 6.1.1 of the ITC, on which OCBC and the Teras Entities rely, states as follows:
[22] The Teras Entities were subsequently struck out as parties as the Teras Entities went into liquidation, and the liquidator did not wish to continue participating in the action.
The decision below
[23] In the proceedings below, the Judge considered five broad issues in finding for OCBC. We briefly summarise the Judge’s decision.
[24] First, the Judge was satisfied that OCBC had proved that the loss was caused by a peril of the seas. OCBC did not bear a burden of showing that there was an aperture that permitted water to enter the Vessel in the first place, and what created it (Judgment at [104]). The Insurers would only have a defence if they were able to demonstrate that the Vessel was a decrepitude (Judgment at [108]). The Judge also found that there was insufficient evidence to demonstrate that the Vessel was inherently unstable or unseaworthy (Judgment at [113]). The Judge preferred the expert evidence of OCBC’s expert, Ms Rosalind Blazejczyk (“Ms Blazejczyk”), that the capsize resulted from an estimated ingress of 450 tonnes of water, although Ms Blazejczyk could not pinpoint the cause of the initial seawater ingress (Judgment at [127]–[129]). Having been satisfied that there was unexpected flooding/seawater ingress in the hull compartments, the Judge found on a balance of probabilities that the capsize was caused by perils of the seas (Judgment at [152]).
[25] Second, the Judge was satisfied that OCBC had proven that the Vessel was a CTL. To discharge its burden of proof, OCBC relied on a number of documents, including correspondence between the Owners and the Insurers, quotations and emails from third parties, and reports on the Vessel’s condition (“CTL Documents”) (see Judgment at [45]). The Insurers objected to the reliance on these documents on the basis of hearsay as the maker of the documents did not testify in court. The Judge was satisfied that the CTL Documents were admissible under s 32(1)(b)(iv) of the Evidence Act 1893 (2020 Rev Ed) (“EA”), also known as the “business records exception” (Judgment at [55]–[59]). Thereafter, the Judge exercised his discretion to cure OCBC’s non-compliance with s 32(4) of the EA in failing to file a notice to admit hearsay evidence, and declined to exercise his discretion under s 32(3) of the EA to exclude the evidence in the interest of justice (Judgment at [60]–[75]). Finally, the Judge found the evidence sufficient to make out a prima facie case of a CTL, and that the Insurers failed to adduce evidence to rebut this prima facie case (Judgment at [76]–[86]).
[26] Third, the Judge found that OCBC did not breach any warranty in the MI. Of relevance to this appeal are two warranties alleged by the Insurers to have been breached:
(a) “all statutory or regulatory requirements whether arising before or during the period covered by this insurance shall be complied with insofar as they relate to the seaworthiness of the vessel” (“Warranty No 1”); and
(b) “all arrangements for moves shall be in accordance with standard operational procedures” (“Warranty No 2”).
For Warranty No 1, the Judge interpreted “statutory or regulatory requirements” as “statutes or regulations (which are subsidiary legislation promulgated by statutes)” (Judgment at [254]). No breach was found on the basis that the instruments raised by the Insurers were neither statutes nor subsidiary legislation (Judgment at [256]). As regards Warranty No 2, the Judge found that it followed from the analysis on Warranty No 1 that there was no breach of Warranty No 2 (Judgment at [265]).
[27] Fourth, the Judge found that there was no breach of the duty of fair presentation. He rejected the Insurers’ allegation that the Teras Entities deliberately failed to disclose that they had “abandoned” Braemar as the MWS because Braemar would not have approved the proposed Tow Voyage and/or that they would only have done so by imposing conditions that the Teras Entities were unwilling to comply with. The evidence merely showed that Braemar presented several “comments for clarification” (Judgment at [198]–[203]). Contrary to the Insurers’ case, the Teras Entities’ failure to disclose the existence of a locking capability in the freeing ports was also not a violation of the prohibition under the International Convention on Load Lines, 1966 and Protocol of 1988, as amended in 2003 (Consolidated Edition, 2005) (“ICLL”). Moreover, the Insurers would not have expected to be informed of a technical matter such as the existence of a locking capability for the freeing ports. The Insurers’ argument concerning the configuration of the freeing ports being different from the “as built” design drawings approved by ABS Houston in May 2016 was dismissed on the basis that it was not pleaded (Judgment at [240]–[241]).
[28] Finally, the Judge rejected the Insurers’ contentions relating to the sufficiency of evidence for the disbursement of the loan from OCBC to TLPL (the borrower) and the breakdown for interest and late charges, which go towards establishing OCBC’s entitlement under the MI. This was because the objections were not pleaded and, in any event, OCBC had discharged its burden of proving the debt of at least US$70m (Judgment at [160]–[169]).
The parties’ cases on appeal
[29] In sum, the Insurers say that the Judge erred because:
[30] OCBC’s case is that the Judge did not err in his reasoning and the appeal should be dismissed in its entirety.
Issues to be determined
[31] Arising from the above, the issues to be determined are:
OCBC did not discharge its burden to prove that the loss was caused by a peril of the seas
Applicable framework to prove loss by perils of the seas
[32] In establishing a claim under an insurance policy, an insured bears the burden of proof to establish loss by an insured peril. Where the insured claims that the loss was occasioned by perils of the seas, the insured must prove so on a balance of probabilities. It is in this context that we consider the available methods of proof.
[33] To set the stage, we begin by briefly considering the scope of “peril of the seas”. Rule 7 of the Rules for Construction of Policy scheduled to the UK’s Marine Insurance Act 1906 (“UK MIA 1906”) stipulates: “The term ‘peril of the seas’ refers only to fortuitous accidents and casualties of the seas. It does not include the ordinary action of winds and waves”. As stated in Arnould: Law of Marine Insurance and Average (Sweet & Maxwell, 21st Ed, 2024) (“Arnould”) at para 23-11, “[i]n order to come within this definition, it is sufficient that the event was in some way fortuitous, it does not have to be shown that it was brought about by extreme weather or any violent action of winds and waves”. There are thus two elements as to what constitutes a peril of the seas: first, it must be a fortuity and not something that is bound to happen, and second, the peril which gives rise to the marine casualty must be “of the seas” (Versloot Dredging BV v HDI Gerling Industrie Versicherung AG [2013] 2 All ER (Comm) 465 (“The DC Merwestone”) at [31] and [34]). The latter element is most commonly satisfied by establishing seawater ingress, but the former element still has to be proved (Howard Bennett, Law of Marine Insurance (Oxford University Press, 2nd Ed, 2006) (“Bennett”) at para 10.26).
[34] In Arnould, the learned authors note that a presumption of loss by perils of the seas may apply, where the vessel is lost without a trace and nothing is known about the prevailing circumstances at the time of her loss, and the vessel is shown to be seaworthy on sailing. However, where there is evidence of an incursion of seawater in a particular part of the vessel the cause of which is uncertain, there is no room for the presumption to apply (at para 20-35):
[35] This is consistent with the framework described in Bennett. Prof Howard Bennett explains that an insured may seek to discharge the burden of proving fortuity either directly or by inference from circumstantial evidence by way of a presumption (Bennett at para 10.26).
(a) First, an insured may directly prove fortuity by establishing, on a balance of probabilities, the precise cause of the seawater ingress and its fortuitous nature. An insured may also prove, on a balance of probabilities, that the cause of the ingress was one of a number of identified possibilities all of which are maritime and fortuitous in nature (Bennett at para 10.27).
(b) Second, an insured may prove fortuity by circumstantial evidence, where the burden of proof cannot be discharged by direct evidence. An insured is entitled to recover for loss by perils of the seas, where a vessel is seaworthy at the time of departure and subsequently sinks in unexplained circumstances (Bennett at para 10.28). Prof Bennett also notes that the question of proof by circumstantial evidence has been raised in circumstances where the presumption cannot arise as the conditions to the presumption are not made out. This is where seaworthiness of the vessel is not proven or where evidence is available as to the circumstances surrounding the sinking of the vessel (at para 10.32).
[36] From the above, we discern two conventional ways in which an insured may prove loss by a peril of the seas on a balance of probabilities.
(a) First, an insured may directly prove that the loss of the vessel was caused by a peril of the seas by putting forward a cause for the loss of the vessel which is attributable to perils of the seas. An insured would have to prove this cause on a balance of probabilities.
(b) Second, an insured may resort to circumstantial proof and rely on a rebuttable presumption that the vessel was lost by perils of the seas, if it can prove that the vessel was seaworthy and that it was lost in wholly unexplained circumstances. In doing so, an insured need not propound a cause for the loss of the vessel.
[37] Where the conditions to the presumption are not made out and the presumption cannot be relied on, an insured may still rely on inferential reasoning and the elimination of possibilities to bolster its case. However, in this situation, an insured must propound an explanation for the loss of the vessel which is attributable to a peril of the seas. The court must be satisfied that this explanation, in light of the elimination of uninsured losses, is rendered more likely than not (or in other words, proven on a balance of probabilities) to be the cause of the loss (see Bennett at para 10.42). This, in our view, constitutes direct proof of a peril of the seas on a balance of probabilities (ie, falling under [36(a)] above). It is strictly not a third method of proving loss by a peril of the seas.
[38] This sets the context against which we analyse the Judge’s decision and the parties’ cases. We note that the above framework is consistent with other English cases, including Lamb Head Shipping Co Ltd v Jennings [1994] 1 Lloyd’s Law Rep 624 (“The Marel”) and The Popi M. At an appropriate juncture, we will elaborate on those cases, and the various conditions applicable to each method of proof.
[39] While the insured strives to establish either methods of proof, the insurer may seek to demonstrate that the loss was due to an uninsured peril such as an uneventful decrepitude of the vessel or inherent characteristics of the vessel. If it is able to do so, it would follow that the loss was not caused by an insured peril or that the presumption has been successfully rebutted. In short, the insurer is entitled to put forward an alternative case theory to explain the cause of the casualty. However, it must not be overlooked that the insurer has no burden to prove the alternative theory (Arnould at para 20-35, The Popi M at 951B–951C). Rejection of the insurer’s alternative theory does not mean that the presumption therefore applies or that the fortuity has been proved.
OCBC’s case below and the Judge’s reasoning
[40] It is important to bear in mind OCBC’s pleaded case that the capsize was caused by a combination of: (a) the flooding of the Vessel’s port aft thruster room and adjacent machinery space; and (b) the effect of wind and waves. OCBC’s own expert, Ms Blazejczyk, explained in her report that “[h]aving established above that wind and waves alone could not have caused the capsize, and that green water on deck was unlikely at the time of the capsize, the most likely cause of the capsize must have been water ingress to the hull of the [Vessel] resulting in loss of buoyancy”. On the stand, she testified that the prevailing wind and waves did not cause the capsize. On this basis, OCBC failed to prove (b) since that was effectively ruled out by its own expert. That would only leave (a) as the propounded cause of the capsize. However, there was no pleading as to the cause of the flooding.
[41] The Judge referred to case law relating to the presumption, viz, The Popi M and Glowrange Ltd v CGU Insurance plc [2001] All ER (D) 339 (Jun) (“Glowrange”) (Judgment at [96]–[99]). However, it is not clear which method of proof the Judge eventually adopted when he concluded that “on a balance of probabilities, the capsize of the Vessel was caused by perils of the seas” (Judgment at [152]). The Judge, after canvassing the authorities, concluded that there was no burden on OCBC to show that there was an aperture that permitted water to enter the ship in the first place, and what created it (Judgment at [104]).
[42] The Judge began by addressing the Insurers’ defence that the Vessel was a “decrepitude” (Judgment at [108]–[120]). But that issue only arises for consideration after OCBC has established the fortuity. In so doing, the Judge had reversed the burden of proof which lies on the insured to first prove its case ie, the fortuity. The Judge erred by equating his rejection of the Insurers’ alternative theory (ie, the Vessel was a “decrepitude”) as an affirmation of OCBC’s primary case. The Judge undertook a comparison between the parties’ experts and preferred the evidence of OCBC’s expert because he found several unexplained inconsistencies in the Insurers’ expert’s evidence (Judgment at [121]–[152]). The fact that the Judge preferred the evidence of OCBC’s expert over that of the Insurers’ expert does not mean that OCBC has therefore proved that the capsize was caused by perils of the sea, as the Judge appeared to have reasoned (see Judgment at [152]). OCBC’s case must be sufficient on its own to make out its claim.
[43] The Judge’s analysis did not appear to draw a distinction between direct proof of a peril of the seas and circumstantial proof in reliance on the presumption. During the hearing of the appeal, counsel for OCBC, Mr Tan Chee Meng SC (“Mr Tan”), acknowledged that the Judge did not appear to have grounded his decision on the presumption, even though Mr Tan at the close of the hearing clarified that OCBC is nonetheless relying on the presumption and that it could be inferred from the Judge’s reasoning that he had implicitly applied the presumption. Counsel for the Insurers, Mr Chan Leng Sun SC, accepted that the Judge might have implicitly applied the presumption even though he contended that the Judge did not have the correct principles in mind. We therefore turn to examine the two established methods of proof in detail, before assessing OCBC’s case on perils of the seas.
A cause to the seawater ingress must be propounded to directly prove fortuity
[44] We first consider direct proof of peril of the seas. It is common ground that the physical cause of the capsize was seawater ingress. This only goes to establish that the peril was of a maritime nature (see [33] above). However, the fortuitous nature of that peril must still be proved. OCBC contends that “[w]here a vessel has been lost by the ingress or entry of seawater, there is no requirement on the insured to prove what occasioned the ingress”, as long as OCBC can demonstrate fortuity. In contrast, the Insurers contend that as a starting point, the search for the proximate cause requires an inquiry as to what had occasioned the entry of seawater. We agree with the Insurers. In our view, case law pertaining to direct proof of perils of the seas requires the insured to propound a cause of the seawater ingress to prove the fortuity.
[45] As highlighted above, the learned authors of Arnould observe at para 20-35 that an assured must prove loss by a peril insured against, and that “if at the conclusion of the evidence the court is left in doubt as to the true reason of the loss, he cannot succeed in his claim” (see [34] above). At para 23-15, the learned authors emphasise that an incursion of seawater does not ipso facto constitute a peril of the seas as it must be shown that the event was fortuitous:
[46] Furthermore, Glowrange emphasised (at [26]) that as a starting point, for a claimant to discharge its burden of proof, a claimant must put forward an explanation for the seawater entry which was intrinsically sufficiently probable and not attributable to the general debility of the vessel.
[47] OCBC disagrees with this proposition and relies on the Privy Council’s decision of Canada Rice Mills, Ltd v Union Marine and General Insurance Company, Ltd [1941] AC 55 (“Canada Rice Mills”), viz, that “[w]here there is an accidental incursion of seawater into a vessel at a part of the vessel, and in a manner, where seawater is not expected to enter in the ordinary course of things, and there is consequent damage to the thing insured, there is prima facie a loss by perils of the sea” (at 68). Read in context, this passage does not assist OCBC’s case. The Privy Council in the very same paragraph went on to detail what may constitute an accidental unexpected ingress of seawater: “some negligent act, such as improper opening of a valve, or a hole made in a pipe by mischance, or it may be that sea water is admitted by stress of weather or some like cause bringing the sea over openings ordinarily not exposed to the sea or, even without stress of weather, by the vessel heeling over owing to some accident, or by the breaking of hatches or other coverings” (at 68–69).
[48] In this light, we do not think that the Privy Council was suggesting that a cause for the seawater ingress need not be propounded. The Privy Council was merely noting that a loss occasioned by an accidental unexpected ingress of seawater would prima facie be loss by perils of the sea, giving examples of such accidental, fortuitous entry. The premise of the proposition, ie, that the seawater ingress was accidental, would still need to be proved by establishing the cause of the seawater ingress.
[49] Next, OCBC cites Popplewell J’s holding in The DC Merwestone at [42] that “[w]here there is a fortuitous ingress of water, it is often the ingress itself, rather than that which renders it fortuitous, which is to be regarded as the proximate cause of the loss, or at least a proximate cause of the loss”. In the same vein, we do not think that Popplewell J was suggesting that no cause for seawater ingress had to be propounded. Instead, Popplewell J merely held that as a matter of causation, provided the cause of the seawater ingress is fortuitous, the loss would prima facie be by perils of the seas. Taking reference to Canada Rice Mills, Popplewell J concluded that the ingress of seawater during a voyage, fortuitously caused by crew negligence, would constitute a peril of the seas (The DC Merwestone at [36]). On the facts, after noting that “a proximate cause of the loss and damage was the ingress of seawater”, Popplewell J found that “the ingress was fortuitous, having been caused by crew negligence at the loading port” (The DC Merwestone at [40]).
[50] This, however, does not detract from the need to prove that the ingress of seawater was “fortuitous”. Earlier in the judgment, Popplewell J held that “[t]he mere fact of ingress of seawater into a vessel is not itself a peril of the seas”, and that “[w]here, therefore, a loss arises by reason of ingress of seawater through the vessel’s hull, the search for proximate cause involves an inquiry into that which occasions the entry of the seawater” (The DC Merwestone at [35]), bearing in mind his earlier observation at [33] that “[t]he fortuity may lie in what causes the hole, or what causes the seawater to reach or enter the hole, or a combination of both”. This is consistent with our reading of Canada Rice Mills, that to prove a peril of the seas, a claimant would still have to establish the fortuitous nature of the seawater ingress by propounding a cause of the ingress.
[51] Lastly, OCBC relies on Venetico Marine SA v International General Insurance Co Ltd [2013] EWHC 3644 (Comm) (“Venetico Marine”). In that case, Andrew Smith J rejected the insurers’ argument that to establish a loss of perils of the seas, the insured “[had] to show on a balance of probabilities ‘that the alleged grounding probably occurred how, when and where they say it did’” (at [289]–[290]). Again, this has to be read in context. The relevant issues of fact in that case were “whether the grounding was fortuitous (and so a peril of the seas) and whether it was a proximate cause of the damage to the vessel” (at [290]). The insured had advanced a primary submission on the circumstances of the grounding, being that the grounding had occurred by about 10am before the vessel used her engine or tried to do so. Smith J held that even on the insurers’ case that the vessel might have been grounded later, the insured would not have failed to discharge its burden if the insured could show in that alternative scenario the grounding was nonetheless fortuitous and a proximate cause of the damage. Smith J concluded in the 10am scenario that the vessel was grounded by operation of the current (at [281]) and that this was a fortuity as it was not an inevitable event (at [283] and [285]). Further, even in the alternative scenario, that the grounding had occurred later and the vessel had unsuccessfully attempted to use its engine (at [286]), the fortuity was that the current carried the vessel aground before repairs were done (at [291]).
[52] In our view, Venetico Marine merely stands for the proposition that where alternative factual scenarios have been advanced, and the court may be left in doubt as to which factual scenario had actually materialised, the insured may still point to the fortuity within the insurer’s alternative factual scenario and prove that even if the insurer’s case were accepted, the loss would still be occasioned by a peril of the seas. It does not purport to dispense with the requirement that a cause has to be advanced to prove the fortuity. This is consistent with Prof Bennett’s observation that an insured need not establish one particular cause, if it proves that the cause of the ingress was one of a number of identified possibilities all of which are maritime and fortuitous in nature (see [35(a)] above, Bennett at para 10.27).
[53] Therefore, having carefully reviewed the authorities relied upon by OCBC, we are satisfied that OCBC was required to put forward a cause for the seawater ingress to directly prove its case of peril of the seas. At the very least, it has to advance a number of identified possibilities all of which were maritime and fortuitous in nature.
The rebuttable presumption of loss by perils of the seas may only be invoked where the vessel is seaworthy on sailing and it is lost in wholly unexplained circumstances
[54] In the alternative, a claimant may establish that the loss was caused by a peril of the seas by circumstantial proof. In this regard, OCBC has relied on the rebuttable presumption that the loss was caused by a peril of the seas even though it is unclear whether the Judge had applied the presumption. OCBC frames its formulation of this rebuttable presumption as such: “[w]here the insured vessel is seaworthy and has been lost by the entry of seawater in circumstances where seawater is not expected to enter in the ordinary course of things, there is a presumption of fact that the vessel has been fortuitously lost by perils of the seas”.
[55] Having examined the cases, we do not agree with OCBC’s formulation. The presumption may only be invoked where the vessel is seaworthy on sailing, and it is lost in wholly unexplained circumstances. While OCBC appears to accept at other parts of its submissions that the rebuttable presumption only arises where there is an “unexplained loss” of a seaworthy vessel, OCBC’s conception of “unexplained loss” appears to encompass situations where the insured simply fails to put forward a plausible explanation, as opposed to cases where the insured is realistically incapable of providing an explanation (due to, for instance, the loss of the vessel at sea). In our view, the latter conception of “unexplained loss” is more consistent with case law. We explain.
The genesis and rationale of the rebuttable presumption
[56] The rebuttable presumption has its genesis in cases where no direct evidence of the loss of the vessel could be adduced. The case of Green v Brown (1743) 93 ER 1126 (“Green v Brown”) is regarded as the progenitor of this presumption (see Bennett at para 10.29). There, a vessel, after setting sail, was never heard of again. The policy excluded loss by capture and seizure, and the underwriters contended that the assured should have adduced direct evidence of the loss in the particular manner declared. The jury was directed that “it would be unreasonable to expect certain evidence of such a loss, as where [everybody] on board is presumed to have drowned; and all that can be required is the best proof the nature of the case admits of”. Seen in this light, the rebuttable presumption was (at least initially) designed to be an evidential crutch for shipowners in an era where ship disappearances were more commonplace and evidence as to the circumstances in which a vessel was lost may not be readily available. Lacking any evidence as to what had occurred to the vessel, it would not be possible for a shipowner to mount a claim on a marine insurance policy. The presumption could, perhaps, be seen as a “risk allocation device that places the risk of loss from unascertainable causes on the insurers” (Bennett at para 10.28).
[57] This rebuttable presumption has been described to be “merely shorthand for the normal inferential conclusion of accidental maritime loss that flows from proof of initial seaworthiness coupled with an absence of further information and the civil law standard of proof on the balance of probabilities” (Bennett at para 10.31). It is made possible by the extensive concept of peril of the seas (Skandia Insurance Co Ltd v Skoljarev (1979) 26 ALR 1 (“Skandia”) at 15). In The DC Merwestone, Popplewell J held at [32] that fortuity may be most easily defined by its antithesis – the cause of the loss should not have been intentional, as where the ship is scuttled, or inevitable, as where there is a rotten hulk or “debilitated” vessel. By rendering the latter unlikely through a finding of seaworthiness and given the improbable nature of the former when nothing has been heard of the crew, the inference to be drawn is that on a balance of probabilities, the ship was lost to a peril of the seas. As the English Court of Appeal in The Marel put it (at 629):
[58] Prof Bennett, with reference to Green v Brown and The Marel, offered a similar analysis as to the true nature of the so-called “presumption”, which we broadly agree with (Bennett at para 10.31):
[59] We thus turn to consider in detail the conditions upon which the presumption of loss by a peril of the seas may be invoked.
The vessel must be proved to be seaworthy on sailing
[60] First, for the presumption to be relied upon, the vessel must be proven to be seaworthy. Seaworthiness is not an absolute concept and should be assessed relative to the nature of the ship, to the particular voyage and even to the particular stage of the voyage (Ever Lucky Shipping Co Ltd v Sunlight Mercantile Ptd Ltd [2003] SGHC 80 at [40]).
[61] One should bear in mind that in this analysis, the “seaworthy” condition is meant to effectively rule out unseaworthiness as the sole proximate cause of the casualty. In other words, if the vessel was proven to be seaworthy, by an elimination of probabilities, it is more likely than not that she had been lost as a result of a peril of the seas. In this light, we agree with the formulation in Glowrange as to the requisite extent of seaworthiness to be proven, “that the relevant condition of the vessel for the purposes of this inference is that it was sufficiently seaworthy to withstand the ordinary action of the wind and waves, that is to say that it was in all relevant respects in a condition better than that of general debility, and not that it was in such a condition as to be able to withstand adverse but unexceptional sea conditions” (at [33]).
The vessel must have been lost in wholly unexplained circumstances
[62] Second, the vessel must have been lost in unexplained circumstances. There is an issue, however, as to what constitutes unexplained circumstances, and whether circumstantial proof by inferential reasoning is available where there is evidence as to the circumstances of the loss of the vessel, eg, that the vessel was lost by seawater ingress, but no cause as to the seawater ingress could be identified by direct evidence on a balance of probabilities.
(1) The content of “unexplained circumstances”
[63] To begin, the rebuttable presumption has been successfully invoked in a number of cases where the ship was lost at sea or the crew has perished, was lost or was otherwise unavailable.
(a) In Green v Brown, the vessel was lost after setting sail and never found again (see [56] above).
(b) In Koster v Reed (1826) 6 B & C 19 359, the ship sailed on a voyage insured with goods on board but did not arrive at her port of destination. A few days after departure, a report was heard that the ship had foundered at sea, but the crew were saved. The defendant contended that given the crew’s apparent survival, it was incumbent on the plaintiff to call them as witnesses (at [20]). The issue was whether there was sufficient evidence to be left to the jury as evidence of a loss by perils of the sea or by barratry. Abbott CJ held that it was not incumbent on the plaintiff to search for the crew, who were assumed to be foreigners, all over Europe (at [22]). In the same vein, Bayley J held that a ship has not been heard of when “no intelligence has been received from persons capable of giving an authentic account; and not that mere rumours have never been heard”, and that it was fair to presume that the ship perished at sea (at [22]).
(c) In Yeung Kong Yung v The Young Shing Investment Co Ltd [1921] HKCU 5, a vessel, the “Luen On”, embarked on a voyage from Hong Kong to Manila in 1917. After departure, she was not heard of again. She was insured against sinking caused by running on rocks or reefs or by damage due to winds or waves. The Hong Kong Full Court noted that “the fact that she has never been heard of raises the well-known presumption of insurance law that she foundered at sea: in other words that she was lost owing to a peril of the seas” (at 25). That said, the court was satisfied, on the meteorological evidence, that the “Luen On” encountered a gale, as described by the captain of another ship operating in the region, which would be particularly dangerous to her (at 27). Therefore, it might not have been necessary for the claimant to rely on the presumption.
[64] By contrast, in three more recent cases, the English courts have held that the presumption could not apply because, at least in part, the loss did not occur in wholly unexplained circumstances.
[65] In The Popi M, the crew survived, and the trial judge could make clear and positive findings as to how the seawater had entered the ship, viz, through a large aperture in the shell plating on the vessel’s port side in way of the engine room, and the manner in which the seawater had flooded other compartments making it inevitable that the ship should sink. Lord Brandon of Oakbrook, delivering the judgment of the House of Lords, noted that the case was “not a case of a ship being lost with all her crew in circumstances when the immediate cause of entry into her of sufficient water to make her sink is unexplained” (at 953A–953B). Therefore, “[i]n the state of knowledge which existed it [was] not surprising to find that the shipowners were strenuously pressed … to specify the perils of the sea on what they relied as having been the proximate cause of creating the aperture in the ship’s shell plating which led to her loss” (at 953B–953C). Accordingly, the shipowners could not rely on the rebuttable presumption of loss by perils of the seas for two reasons. First, there was no finding that the ship was seaworthy. Second, “the loss did not occur in unexplained circumstances: on the contrary, the reasons why she sank, apart from the cause of the fatal aperture itself, were as clear as they could possibly have been” (at 953D–953E).
[66] In The Marel, a bump was felt on board the ship sufficient to throw the crew off balance, and immediately afterwards, large volumes of seawater flooded the engine room on the floor plate level in the vicinity of the high sea chest (at 624). On the known facts, the trial judge had found it wholly improbable that the entry of seawater into the engine room was due to a collision with a container, rejecting the plaintiff’s primary case (at 628). The trial judge therefore was not satisfied on a balance of probabilities that the sinking of the ship in The Marel was caused by perils of the sea. On appeal to the English Court of Appeal, the plaintiff relied on “the presumption or inference that the Court makes when a ship has disappeared at sea and it is shown that she was seaworthy when she set out on her last voyage, that her loss was by ‘perils of the sea’.” The plaintiff submitted that even though “the Marel did not simply vanish without a trace, and there is a certain amount of evidence as to the circumstances of her sinking, nonetheless the presumption applies insofar as the evidence available does not provide a complete explanation of how she came to sink” (at 629). Dillon LJ did not accept the submission, holding that it was inconsistent with previous propositions as to the onus of proof, because the plaintiff’s stance would mean that if a ship was seaworthy when she set out and is lost, the loss must be held to have been due to the perils of the seas, unless it was affirmatively established by the underwriters otherwise. Dillon LJ was also not persuaded that the presumption could apply as the known circumstances of the sinking eliminated a number of possible perils of the seas, such that there was “no room for the presumption”. It was for the plaintiffs to prove their case (at 629).
[67] In Brownsville Holdings Ltd v Adamjee Insurance Co Ltd [2000] 2 All ER (Comm) 803, Aikens J held at [18] that “the presumption only operates when the loss of the vessel is unexplained”, explaining with reference to The Marel that the presumption did not operate in favour of the claimants “when they have not been able to put forward a plausible explanation for the loss or their explanations have been rejected by the court as too improbable for belief, so that the loss is ‘unexplained’ in that sense”. Instead, the presumption only applies where a seaworthy ship disappears in “wholly unexplained circumstances”. On the facts, the presumption was not applicable as “the crew were rescued and have told their story in evidence”.
[68] This view appears to be shared by commentators. The learned authors of Arnould express the view that “[i]t is only where the loss is ‘wholly unexplained’ that such presumptions may be invoked. If there is direct evidence from those on board regarding the casualty, the loss is not wholly unexplained” (Arnould at para 22-52, fn 388). Likewise, Professor Bennett notes that the vessel must be proven seaworthy and is lost without trace, or alternatively, in waters where physical access is impossible and the master and crew must be presumed dead (Bennett at para 10.31).
(2) Availability of inferential reasoning where the loss did not happen in wholly unexplained circumstances
[69] We, however, note that a more permissive approach in relation to proof by inferential reasoning by elimination appears to have been articulated in several cases. This is such that a claimant may be able to eliminate unseaworthiness and scuttling as causes and therein establish perils of the seas, even where the loss did not happen in wholly unexplained circumstances, eg, where there is direct evidence of the circumstances surrounding the loss of a ship. While OCBC’s case was framed as a “presumption of fact” and not as a matter of proof by inferential reasoning, the central thrust of its submission, ie, a seaworthy ship which capsized by entry of seawater should be found to be fortuitously lost by perils of the seas, engages with this broader issue of inferential reasoning. We therefore consider this question for completeness.
[70] In Glowrange, the English High Court had the opportunity to consider the question of proof of loss in light of The Popi M and The Marel. The court held that while there may be evidence as to how the vessel came to sink, such as the failure of the hull at a particular place and in particular sea conditions, the loss might still be established to be caused by perils of the seas if the assured can prove that the vessel’s condition was not so debilitated that it could not withstand the ordinary action of wind and waves in calm conditions (at [18]). The court observed that proof of loss by perils of the seas did not fall into some special category of forensic exercise, and set out two propositions.
[71] It would thus appear that the court in Glowrange was of the view that even if the vessel was not lost in wholly unexplained circumstances, proof by inferential reasoning by elimination may still be available to a claimant, as an insured may eliminate all suggested uninsured causes to make out its case that the loss was occasioned by perils of the seas.
[72] At [26], the court in Glowrange construed Lord Brandon’s judgment in The Popi M. There, the House of Lords held that it was impermissible for the court to accept a wholly improbable reason as proven on a balance of probabilities simply by a method of elimination. The House of Lords rejected Sherlock Holmes’ celebrated investigation theory, that once “[y]ou have eliminated the impossible, whatever remains, however improbable, must be the truth”, on three grounds (at 955H–956E). First, the judge was not bound to make a finding one way or the other and it was open to find a third alternative that the burden of proof has not been discharged. Second, the dictum can only apply when all relevant facts are known and all possible explanations except a single extremely improbable one can properly be eliminated. Third, a finding that an extremely improbable event is nonetheless more likely to have occurred than not, does not accord with common sense. The House of Lords held that the trial judge should not have accepted the shipowners’ wholly improbable case of perils of the sea simply because the underwriters’ theory was ruled as impossible (at 955D–955H). The court was left in doubt as to the true cause of the loss, as “[t]he shipowners failed to establish the ship was seaworthy, and they only put forward an extremely improbable cause of the loss” (at 956F–956G).
[73] In this light, the court in Glowrange opined that in The Popi M, had the vessel been found seaworthy, the insured would have established that it was to be inferred that the seawater entry must on a balance of probabilities have been due to perils of the seas without the need to advance any particular explanation (at [26]).
[74] The court in Glowrange at [31] also considered The Marel, and its reference (at 632) to a passage from The Gloria (1936) 54 Ll L Rep 35 which stated that “[i]f, however, the evidence is such that the court giving full weight to the consideration that scuttling is a crime is not satisfied that the ship was scuttled but finds that the probability she was is equal to the probability that her loss was fortuitous the plaintiffs will fail”. This was taken to indicate that if there is “neither a specific and cogent explanation for a loss by perils of the seas nor any possibility of real substance that the vessel could have been scuttled”, the court is “entitled to infer that the loss happened in a way that is unknown, but by perils of the seas, by looking at the whole of the evidence before it, including in particular, evidence that the vessel was in all relevant respects seaworthy at the start of the voyage” (Glowrange at [32]).
[75] However, the court accepted that there may be cases in which primary evidence is available to suggest from where seawater entered a vessel but does not indicate why the entry occurred. In those cases, the insured will have to advance a cogent explanation for the seawater entry that is strong enough to displace all other uneliminated but uninsured perils. The court construed The Popi M and The Marel as cases in which the omission to do so had left the court in such doubt that it was unable to infer that the loss was more probably caused by perils of the seas than an uninsured peril (Glowrange at [34]).
[76] The court was also of the view that The Popi M did not disturb the well-established approach to unexplained losses exemplified by Anderson v Morice (1874) LR 10 CP 58 (“Anderson v Morice”) and Ajum Goolam Hossen v Union Marine Insurance Co Ltd [1901] AC 362 (“Ajum Goolam”) (Glowrange at [26]) though we should add that both cases did not strictly deal with the issue as to whether the loss of the vessels were caused by a peril of the seas. Instead, both cases concerned claims for loss of insured cargo. In Anderson v Morice, the main issue was whether the plaintiff had an insurable interest in the cargo since the vessel sank prior to the completion of the loading operation whereas in Ajum Goolam, the dispute was whether the vessel was seaworthy given that she had developed an unusual list before she left the harbour. In both cases, there appeared to be some evidence available as to the circumstances of the loss.
[77] We also note that OCBC has relied on Skandia. In Skandia, the High Court of Australia appeared to have engaged in inferential reasoning, where the loss was “unexplained” in that the plaintiffs did not propound a cause of the seawater ingress. The ship sank a few hours after leaving port due to the rapid entry of seawater into the engine room, and notably, the crew survived to give evidence as to the circumstances of the sinking (at 5–6). The plaintiffs’ case was that the ship was seaworthy and that the sinking was due to some unknown but fortuitous cause, relying on the presumption or inference that the cause of the sinking was a peril of the seas (at 6).
[78] The trial judge was unable to make a finding as to the cause of the seawater ingress and did not accept the evidence of a marine surveyor called by the plaintiffs, that the seawater ingress was caused by a burst pipe. Instead, the trial judge applied the presumption that if a ship which is seaworthy sinks in smooth water and there is no other evidence as to the cause of the loss, the casualty is attributable to a peril of the seas (at 7). The High Court of Australia was satisfied that the trial judge’s finding did not depend upon the onus of proof, as it amounted to “a positive rejection of the appellant’s case and a finding that the [ship] was seaworthy when she left port” (at 17). On this premise, the High Court held that “unseaworthiness was excluded as the cause of the loss and the court, quite correctly, inferred that the loss was attributable to a peril of the sea, though it was unable to identify that peril” (at 17). The High Court was satisfied that “[i]t was not a case in which the evidence as to competing causes of loss was evenly balanced, leaving the court in doubt upon the issue whether the loss was attributable to a peril of the sea or to some other cause such as the unseaworthiness of the vessel” (at 17).
[79] Tying the strands of case law together, in our view, the presumption, which dispenses with a need for the claimant to propound any reason at all for the loss of the vessel, should only be invoked where the vessel is lost in “wholly unexplained circumstances”. This requires that the claimant must not be in a position to investigate or determine the cause of the loss. In other words, the claimant must be unable to identify the actual cause of the loss due to a deficit of direct evidence. As a corollary, where there is sufficient direct evidence which may suggest how the vessel was lost, or where the claimant could investigate the cause of the incident but did not do so, the claimant is not entitled to invoke the presumption. The claimant may nonetheless rely on inferential reasoning by elimination to directly prove a fortuity insofar as the claimant propounds a cause and relies on circumstantial evidence as support to prove the cause on a balance of probabilities.
[80] To begin, we note that Glowrange suggested that the need to condescend to particulars should be understood in the context of the unusual facts in The Popi M and that had seaworthiness been found, there would not have been a need to advance any particular explanation (at [26]). In other words, The Popi M did not truly rule out all possible explanations as the vessel was not found to be seaworthy. However, this very argument was raised in The Marel and rejected. The plaintiffs in The Marel submitted that two possible alternative causes of casualty were open to the court in The Popi M; in contrast, in The Marel, unseaworthiness was ruled out as a cause of the incursion of seawater into the engine room, and scuttling was not put forward as an affirmative case (The Marel at 629–631). The plaintiffs thus contended that the judge was bound to accept the one possibility open to him – loss by perils of the seas (The Marel at 632). This was rejected by the English Court of Appeal in The Marel, which held that the third reason articulated in The Popi M – that it did not accord with common sense to find that an extremely improbable event was nonetheless more likely to have occurred than not – squarely applied.
[81] The position in The Marel comports with the Appellate Division of the High Court’s understanding of The Popi M as articulated in Chubb Insurance Singapore Ltd v Sizer Metals Pte Ltd [2023] 1 SLR 1553. The court held (at [70]):
[82] In this light, we think that the inquiry on the availability of direct evidence as to the actual circumstances of the loss is directed to ascertain whether a case of perils of the sea can be said to be improbable. The available evidence may rule out potential causes relating to a peril of the seas. For instance, in The Marel, the available evidence ruled out a number of perils of the seas – the ship was not overwhelmed by exceptionally bad weather or run down and sunk by another vessel. The possibility of collision with an unidentified object was shown to be wholly improbable and very nearly impossible (at 629). The case of loss by perils of the sea was thus shown to have been “wholly improbable and very nearly impossible” (at 633). In contrast, where a vessel is lost in wholly unexplained circumstances, there would be no basis to rule out possible causes of perils of the seas to render a case on perils of the seas improbable. Thus, the presumption properly comes into play.
[83] The same proscription in The Popi M and The Marel against the acceptance of a wholly improbable cause should, in the same vein, apply where a claimant omits to advance a cause in relation to perils of the sea, despite known circumstances enabling the claimant to do so. Such an omission leaves the court unable to assess the probability of the claimant’s case of loss by a peril of the seas. It may even suggest that any theory put forward by the claimant is improbable.
[84] Further, we bear in mind that a finding of seaworthiness does not fully exclude the possibility that the vessel could have been a debility and that the loss was inevitable due to the vessel’s condition. A finding of seaworthiness is made on a balance of probabilities, premised on the available evidence. The inferential value arising from a finding of seaworthiness may vary from case to case depending on the strength of the evidence (see Glowrange at [33]). But even where there is strong evidence supporting a finding of seaworthiness, it merely renders inevitable loss by debility improbable as it “remains possible that a seaworthy vessel might be lost by reason of a lack of fitness that it was not reasonable to expect” (Bennett at para 10.31). There would not be a basis to prefer one improbable explanation (of perils of the sea) over another improbable one (of inevitable loss by debility).
[85] In this light, we emphasise that it must be incumbent on a claimant to advance a cause where there is sufficient evidence on the actual circumstances of the loss. This is broadly consistent with the acceptance in Glowrange, that where there may be primary evidence which suggests from where seawater entered a vessel, a cogent explanation for the seawater entry would have to be propounded (at [34]), even though we would phrase it differently and require an explanation so long as sufficient direct evidence is available as to the actual circumstances of the loss. The failure to advance a cause despite sufficient direct evidence being available may leave the court in such doubt that it is unable to make a finding by elimination that the loss was caused by perils of the sea. This would mean that inferential reasoning is available to the extent that it supports direct proof of loss by perils of the seas, by supporting a finding that a particular cause is proven on a balance of probabilities.
[86] Further, it must be borne in mind that the presumption is to assist an insured shipowner who is unable to identify the actual cause of the loss to discharge its burden of proof. Where a vessel is lost at sea without a trace, the shipowner would not have the means to prove the cause of the loss. In such a situation, the law allows the shipowner to rely on the presumption that the loss was due to “perils of the sea” provided the vessel is otherwise seaworthy, thus dispensing with any need to advance a cause as to the seawater ingress. However, where the loss is capable of being explained, the presumption cannot be invoked, and it is plainly insufficient for the shipowner to claim that the cause of the seawater ingress is unknown.
[87] A comparison can perhaps be drawn to the maxim of res ipsa loquitor in the tort of negligence. This maxim has been construed to be a “convenient label to apply to a set of circumstances in which the claimant proves a case so as to call for a rebuttal from the defendant, without having to allege and prove any specific act or omission on the part of the defendant” (Andrew Tettenborn et al, Clerk & Lindsell on Torts (Sweet & Maxwell, 24th Ed, 2023) (“Clerk & Lindsell”) at para 7-208). Res ipsa loquitor applies where: (a) the occurrence is such that it would not have happened without negligence; and (b) the thing that inflicted damage was under the sole management and control of the defendant. However, there is a third requirement, that the cause of the occurrence is unknown to a claimant (see Clerk & Lindsell at paras 7-208 and 7-211; see Grace Electrical Engineering Pte Ltd v Te Deum Engineering Pte Ltd [2018] 1 SLR 76 (“Grace Electrical”) at [39]). As Lord Nomand in Barkway v South Wales Transport Co Ltd [1950] 1 All ER 392 described, the maxim is “no more than a rule of evidence affecting onus. It is based on commonsense, and its purpose is to enable justice to be done when the facts bearing on causation and on the care exercised by the defendant are at the outset unknown to the plaintiff and are or ought to be within the knowledge of the defendant” [emphasis added] (at 399).
[88] This third requirement, that the cause of the occurrence is unknown to the claimant, is met where the precise cause of an incident had not been established on a balance of probabilities, even where there is some evidence available as to the circumstances of the incident (Grace Electrical at [73]–[83]). Despite this formulation, res ipsa loquitor applies only in cases where “there is genuine difficulty with establishing the cause of the incident and not in cases where, merely by reason of the way the case was run, there was no evidence on the relevant issues before the court” (Grace Electrical at [84]). In other words, “the rule is a practical outworking of the burden of proof in cases where there are real difficulties in establishing what in fact happened, and not a means by which to overcome the shortcomings in the evidence arising only from the failure of the plaintiff to prove his case in the appropriate way” (Grace Electrical at [84]).
[89] In our view, the same consideration should apply to the availability of the presumption. The presumption should not be used to overcome shortcomings in the evidence arising from the manner in which a claimant has chosen to prove his case. Instead, the claimant must genuinely be unable to establish the cause of the incident, due to a deficit of direct evidence. This would ensure that the burden of proof is squarely on the claimant where the claimant is in a position to establish what had in fact happened, while fairly enabling justice to be done when the circumstances are unknown to the parties.
[90] Finally, we note the UK Supreme Court had in Global Process Systems Inc v Syarikat Takaful Malaysia Bhd [2012] 1 All ER (Comm) 111 casted some doubt on Skandia. The court observed at [73], in relation to the finding in Skandia that there was an inference of some unidentified accident or fortuitous event, that “since The Popi M, more attention might have been given, in this jurisdiction at all events, to a finding that no cause had been shown to be more probable than not”. In our respectful view, this observation is consistent with our elucidation of the law. In Skandia, given the existence and availability of significant direct evidence on the circumstances of the loss, eg the survival of crew, surrounding weather conditions, rapid entry of seawater into the engine room, it could be said that the presumption ought to have been unavailable and if inferential reasoning was to be engaged, it was perhaps incumbent on the plaintiffs to advance a cause for the loss of perils of the sea and prove that cause on a balance of probabilities.
[91] In the same vein, we note that to the extent that Anderson v Morice and Ajum Goolam may stand for the proposition that the court may, on inferential reasoning, conclude that the loss of a ship (as opposed to cargo, which was the subject of the claims in those cases) could have been caused by “some extraordinary though invisible and unascertained accident of the seas” (Anderson v Morice at 68), this proposition is in conflict with the salutary caution in The Popi M and The Marel against finding that an improbable cause is nonetheless proven by elimination on a balance of probabilities. To reiterate our view, a cause should be propounded such that the court can conclude that that cause is proven on a balance of probabilities.
Summary of the applicable law
[92] To recapitulate, when a ship which is otherwise seaworthy is lost at sea in circumstances which are wholly unexplained, the law presumes that the loss was caused by an insured peril, ie, the perils of the sea. The italicised words form the two conditions to invoke the presumption. The “seaworthy” condition if satisfied would effectively rule out unseaworthiness as the sole proximate cause of the casualty. The “wholly unexplained” condition applies when the shipowner is not in a position to investigate or determine the cause of the incident, in this case, the cause of the seawater ingress.
[93] It must be borne in mind that the presumption is to assist an insured shipowner who is unable to identify the actual cause of the loss to discharge its burden of proof. Where a vessel is lost at sea, the shipowner would not have the means to prove the cause of the loss. In such a situation, the law allows the shipowner to rely on the presumption that the loss was due to “perils of the sea” provided the vessel is otherwise seaworthy. However, where the loss is capable of being explained as there is sufficient direct evidence, the presumption cannot be invoked, and it is plainly insufficient for the shipowner to claim that the cause of the seawater ingress is unknown. It is then incumbent on the shipowner to directly prove perils of the seas by advancing a cause as to the seawater ingress and proving the cause on a balance of probabilities, with the available evidence.
On the facts, OCBC has not proved that the Vessel was lost by a peril of the seas
OCBC did not propound a cause of the seawater ingress
[94] As a matter of direct proof of loss by a peril of the seas, OCBC must propound a cause of the seawater ingress to prove the fortuity (see [44]–[53] above). We are of the respectful view that the Judge could not have concluded solely from OCBC’s expert evidence that the capsize was caused by a peril of the seas. This is because OCBC did not advance a positive cause of the seawater ingress, which is necessary to prove the element of fortuity.
[95] The Judge preferred the evidence of OCBC’s expert, Ms Blazejczyk, who testified that there was unexpected flooding or seawater ingress in the hull compartments (Judgment at [152]). Ms Blazejczyk estimated that for the Vessel to have attained the position it did at 4.25pm on 5 June 2018 as exhibited in a photograph, there would have been around 450 tonnes of seawater ingress at that time (Judgment at [127]). However, Ms Blazejczyk did not purport to offer an opinion of which route of seawater ingress is more probable than others. She did not testify as to the likely causes of seawater ingress based on the actual condition of the Vessel as surveyed in the Dive Inspection Report; instead, she merely purported to theorise three feasible routes of seawater ingress.
[96] Conversely, the Insurers had contended, on the evidence of their experts, Mr Hu Zhigong (“Mr Hu”) and Cpt Nicolas James Allan White (“Cpt White”), that (Judgment at [135]–[136]):
[97] The Judge rejected the evidence of the Insurers’ expert, Mr Hu, as flawed due to several unexplained inconsistencies in his theory (Judgment at [140]). As Cpt White’s evidence on the capsize was largely built on Mr Hu’s thesis, the Judge likewise rejected Cpt White’s evidence (Judgment at [148]). Having accepted Ms Blazejczyk’s evidence and rejected the evidence of the Insurers’ experts, the Judge found on a balance of probabilities that the capsize of the Vessel was caused by perils of the sea (Judgment at [152]).

The judgment continues past this page. Read the rest at eLitigation.

The insured did not prove the loss was caused by perils of the seas

Laws this judgment relies on

Grounds of decision

Introduction

Background facts

The Tow Voyage

Events after the Vessel’s capsize

The decision below

The parties’ cases on appeal

Issues to be determined

OCBC did not discharge its burden to prove that the loss was caused by a peril of the seas

Applicable framework to prove loss by perils of the seas

OCBC’s case below and the Judge’s reasoning

A cause to the seawater ingress must be propounded to directly prove fortuity

The genesis and rationale of the rebuttable presumption

The vessel must be proved to be seaworthy on sailing

The vessel must have been lost in wholly unexplained circumstances

(1) The content of “unexplained circumstances”

(2) Availability of inferential reasoning where the loss did not happen in wholly unexplained circumstances

Summary of the applicable law

On the facts, OCBC has not proved that the Vessel was lost by a peril of the seas

OCBC did not propound a cause of the seawater ingress