Bunge SA v ADM DO Brasil Ltd & Ors

This is an appeal brought with permission of Aikens J granted pursuant to section 69 of the Arbitration Act 1996 against eight arbitration awards issued by maritime arbitrators in relation to a dispute concerning the shipment of allegedly dangerous cargo. The cargo in question was 44,337.515 tonnes of Brazilian soyabean meal pellets, “SBMP”, loaded on board the “Darya Radhe” at Paranagua by nine shippers between 20 and 27 January 2004. The arbitrators found that there had been introduced with this cargo during loading less than 20 and probably no more than 14 live rats. Discovery of the rats during loading was said by Bunge SA, (“Bunge”), who were the time charterers of the vessel, to have been responsible for their incurring extraordinary expenditure and delay in dealing with the matter in an appropriate way. Their loss was put at in excess of US$2 million. Bunge say that SBMP loaded with accompanying rats is a dangerous cargo. Bunge accordingly brought claims for damages against the nine shippers to each of which it had issued at least one bill of lading.

There were in fact thirty bills of lading in all. Each named the port of discharge as “Bandar Imam Khomeini or Bandar Abbas, Persian Gulf – Iran”. Each bill of lading incorporated the Hague Rules. The Hague Rules provide, by Article IV Rule 6:

“6.

Goods of an inflammable, explosive or dangerous nature to the shipment whereof the carrier, master or agent of the carrier, has not consented, with knowledge of their nature and character, may at any time before discharge be landed at any place or destroyed or rendered innocuous by the carrier without compensation, and the shipper of such goods shall be liable for all damages and expenses directly or indirectly arising out of or resulting from such shipment.

If any such goods shipped with such knowledge and consent shall become a danger to the ship or cargo, they may in like manner be landed at any place or destroyed or rendered innocuous by the carrier without liability on the part of the carrier except to general average, if any.”

It was said by Bunge in the arbitrations that the shippers were liable for the losses which they had incurred as damages and expenses directly or indirectly arising out of or resulting from the shipment of the cargo. In the alternative, Bunge relied upon the term classically regarded as implied into a contract of affreightment, that the shipper of goods will not ship goods of a dangerous character of which the carrier could not by reasonable diligence have become aware before loading.

The arbitrations were heard together, with sensible co-operation as to representation and evidence. Two arbitrators, Mr David Farrington and Mr Christopher Moss, were common to all nine references. In fact there were ten references but one did not proceed to an award and can be ignored. In seven of the references the third member of the tribunal was Mr Anthony Scott. In the remaining two it was Mr Alan Burbidge and Professor Charles Debattista respectively. The arbitrators issued one set of Reasons which were common to all nine references. Bunge has not appealed against two of the awards – this is of no relevance.

The principal contest at the hearing of the arbitrations, which occupied seven days, concerned the factual question whether, as Bunge contended, the rats were introduced with the cargo or whether rather, as the shippers contended, the rats were present on board before loading commenced or had boarded the vessel via inadequately protected mooring ropes and gangways. On this issue the arbitrators considered a considerable body of both factual and expert evidence. In the light thereof the arbitrators, with Mr Burbidge and Mr Scott dissenting, found that the rats had entered with the cargo, and against that finding of fact there is naturally no appeal.

The arbitrators unanimously concluded that Bunge’s claims against each shipper failed, for two independent reasons, each of which is fatal to the claim. First, they concluded that Bunge was unable to prove that any of the shippers, who were wholly independent of each other, had been responsible for introducing even a single rat into the vessel’s holds. There were, as the arbitrators neatly put it, by reference to the number of parcels loaded, “more shippers than rats” and Bunge was unable to show which shipper or shippers was or were responsible. Secondly, the arbitrators concluded that in this case a cargo “loaded with a rat” was not a dangerous cargo. They came to that conclusion in the light of findings of fact which included:

In relation to this last finding the arbitrators observed that it was not clear whether this outcome was the result of the cargo being clean or whether it was because of the arrangement which Bunge reached with the receivers. There is no finding as to what those arrangements were. After successful fumigation at Paranagua the vessel had deviated to Lisbon from her expected voyage via the Cape where, at a facility in which Bunge had an interest, some of the cargo had been discharged and re-inspected and the entire cargo had again been fumigated, notwithstanding there was no basis for thinking that any live rats remained in it. Since the claim failed and the arbitrators had no need to decide quantum issues, they had no need to make a finding about this incident. However they indicated that had they needed to make a finding about it, they would have concluded that the main reason for deviating to Lisbon was a problem with the quality of the vessel’s bunkers, and that diversion to Lisbon simply to conduct these cargo operations was entirely unjustified. The arbitrators did however observe that “the consequences of the vessel arriving in Iran with live rats were of such magnitude that it was difficult to criticise a decision to re-fumigate”. They also observed that it had to be recognised that Bunge has great experience in this field and that it had to be assumed that it would not have incurred these significant expenses without genuine and justifiable concern at the time that there was a continuing risk of live rats in the cargo. Nonetheless the arbitrators concluded that this risk was in fact non-existent. Bunge included a modest claim, US$4,485.00, for the cost of representation at the discharge port in Iran “to safeguard the interests of the vessel”. In relation to this the arbitrators said, at paragraph 129 of their Reasons:

“Experience proves that having the right person on the ground in such circumstances can be a crucial factor in avoiding potential disasters.”

The upshot is that despite the arbitrators’ observation that it was not clear whether non-rejection of the cargo was a consequence of the cargo being clean or because of the arrangements which Bunge reached with the Receivers (Reasons paragraph 77) there are clear findings that the fumigation at Paranagua was successful and that the cargo was discharged at Bandar Khomeini without incident. There is no basis upon which it can plausibly be asserted that, but for the unspecified arrangements with the receivers, the cargo would have been rejected in Iran, still less that the receivers would have been justified in so doing.

Bunge sought leave to appeal on the following points of law:

“(1)

does the presence of between 14 and 20 rats (or any lesser number of rats for which any individual Shipper is liable) in the cargo render the cargo either physically or legally ‘dangerous’?

(2)

if so a further issue may then arise, namely: ‘what is the correct legal test for establishing whether any particular shipper is liable?’, in particular:

(i)

is Bunge required to show [that] each parcel of cargo loaded by a shipper contained a rat or is it enough for Bunge to show that a particular shipper loaded rats with the cargo? And

(ii)

does Bunge need to show that on the balance of probabilities the rats were distributed evenly throughout the cargoes or does Bunge need to show that there was no possibility that the rats were distributed unequally throughout the cargoes?”

In granting permission to appeal Aikens J suggested refinements to the points of law. As redefined, the questions are:

“3.

… ‘Does the presence of between 14 and 20 rats in the cargo loaded on board the vessel MV “Darya Radhe” at Paranagua by one or more of the shippers, render some or all of the cargo loaded “dangerous” for the purposes of the terms of (a) the charterparty incorporated into the Bills of Lading and/or the common law and/or (b) Article IV Rule 6 of the Hague Rules?’

4.

… ‘On the assumption that the answer to the question set out above is “yes”, has the tribunal adopted the correct test on what the claimant carriers have to prove to make one or more of the shippers of cargo at Paranagua liable for (a) damages at common law; and/or (b) all damages and expenses directly or indirectly arising out of or resulting from such shipment in accordance with Article IV Rule 6 of the Hague Rules?’ ”

The parties did not consider that this reformulation of the first issue posed the correct question. The question which they preferred to address was whether a cargo (of SBMP) loaded with a rat, alive or dead, is a dangerous cargo. It is certainly the case that any formulation of the question must in principle focus on whether the cargo loaded by any one shipper was itself dangerous so as to generate a liability in that shipper, irrespective of the position of other shippers. The question for me at this stage is whether, on the basis of the facts found, the arbitrators came to a correct conclusion that, even had it been possible to show that one or more identified shippers was or were responsible for the introduction of the rats, or any one of them, nonetheless that would have involved no breach of contract by the shipper or shippers concerned, nor would it have generated a right to recovery under Article IV Rule 6 of the Hague Rules. The arbitrators’ conclusion was of course that even had it been possible to demonstrate that one particular shipper was responsible for the introduction of all of the rats, nonetheless that shipper would not have been liable to Bunge either at common law or pursuant to the Hague Rules. If that conclusion is correct, alternative hypotheses involving fewer rats or even a single rat do not of course need to be considered. The second question only arises if the arbitrators’ conclusion on dangerous cargo is wrong. In that event the question arises whether the arbitrators erred in their approach to how Bunge was required to satisfy the burden of proof which it assumed in bringing its claims.

Before addressing the legal issues I must first set out the facts as found by the arbitrators in a little more detail. Bunge was both the time charterer of the vessel and the purchaser, on FOB terms, via an intermediary, of the SBMP. Each of the shippers had sold on FOB terms to Ceval International Limited from whom Bunge purchased on like terms. Although the arbitrators make no finding I infer that Bunge had sold the SBMP to Iranian interests on CIF or similar terms.

The SBMP cargo was loaded into Holds Nos. 1, 2, 4, 6 and 7. The cargo was loaded into the vessel between 20 and 27 January 2004 from the AGTL Cotriguacu and Silo Horizontal terminals in the Paranagua Grain Export Corridor. The Export Corridor consists of twenty-one silos located in the port area with a total storage capacity of 760,000 tonnes of agricultural commodities. The silos are connected by means of conveyor belts to four loading spouts installed on the loading wall.

Bunge was itself the shipper of a cargo of 26,600 tonnes of maize which was also loaded on board the vessel at Paranagua between 20 and 22 January from the Cargill, CBL and Centro SUL silos, likewise in the Export Corridor. The maize cargo was loaded into Holds Nos. 3 and 5.

The procedure employed for loading both the SBMP and the maize cargoes was that parcels were released from the various cargo stores onto covered conveyor belts. These took the cargo up to ship loaders which then dropped the product into the vessel’s holds through a loading spout.

The arbitrators’ further essential findings of fact emerge from the following paragraphs of their Reasons:

“11.

Bunge’s case – at least, as it was advanced prior to the oral hearing – was that ‘numerous’ rats were found in the SBMP; in the holds and on deck adjacent to the holds after the commencement but before the completion of loading. It maintained that all these numerous live rats had been loaded onto the vessel with the SBMP cargoes. Shortly after the first record of sightings of rats was made on 24 January 2004, fumigation had been ordered by the Brazilian Ministry of Agriculture which had been certified the ‘presença de pragas’ (‘presence of plague’) on the vessel – which Bunge suggested was probably a reference to the rats. Fumigation of the holds containing SBMP was ordered by the Ministry of Agriculture. It seemed to us to be clear by the close of the hearing that the fumigation was entirely routine – and indeed was a requirement of the sale contract with the Iranian receivers. No evidence of rat infestation was noted either in the maize cargo or the holds into which it was loaded.

12.

In any event, by the conclusion of the evidence at the hearing it had become apparent that a total of no more than eleven rats had been captured, nine being found in the holds containing the SBMP cargoes and two on the main deck.

13.

The Master insisted on clausing the Mate’s receipts for the SBMP with a reference to ‘the discovery of live vermin’. Despite the fumigation of the cargo (on 27 January 2004) Bunge advised the Shippers on 28/29 January that they were rejecting the SBMP cargoes due to the presence of rats and they requested that the SBMP be discharged and replaced with sound cargoes. The vessel was moved to the inner anchorage while arrangements were made to discharge her.

14.

The clausing of the Mate’s receipts led to seven of the ten Shippers making applications to the Paranagua Civil Court which resulted in them obtaining an Order of the Court requiring the Owners to issue clean bills of lading. The Master was then obliged to sign clean bills of lading and Bunge had no alternative but to carry the allegedly infested SBMP cargoes under protest.

15.

The vessel sailed from Paranagua on 19 February 2004. Bunge maintained that as a result of its concerns that the cargo might be rejected on its arrival in Iran it had arranged for the vessel to proceed to a terminal at Lisbon in which it had an interest where the cargo could be re-inspected and for her then to proceed through the Mediterranean and the Suez Canal instead of by the route originally intended via the Cape of Good Hope.

16.

The vessel was inspected at Lisbon between 12 and 17 March. She arrived at Bandar Imam Khomeini, Iran on 9 April and discharged between 10 and 26 April without incident.

…

19.

This dispute had already generated lengthy and involved legal proceedings both in the Courts in Brazil and in this arbitration by the time that it came before us for determination of the substantive issues. However, although we were faced with a substantial volume of both factual and expert evidence, the principal areas of enquiry facing us at this stage were relatively straightforward. The first was as to the extent of the infestation and how it occurred. Bunge’s case was that during the loading operation, live rats were loaded onto the vessel with the SBMP cargoes. The Shippers collectively maintained that the rats were either on board before loading commenced or had boarded via inadequately protected mooring ropes and gangways.

20.

A subsidiary issue of fact was whether, if the rats had come on board the vessel with the cargo, were there rats present in every shipper’s cargo – or only some of them, and if so which? Bunge’s case was that each and every one of the SBMP parcels was infested. The expert evidence focused on the question of which of the two alternative explanations for the presence of the rats on board was the more likely in the light of the circumstances ashore, the behaviour of rats generally and such evidence as there was of the steps taken by the vessel to prevent rats boarding.

…

27.

At this point it is necessary to refer to the order in which the various cargoes were loaded. Counsel for Bunge helpfully prepared the following table (Footnote: ¹) of loading times by reference to the cargoes of each shipper which is attached to the Reasons as ‘Appendix A’. The precise timings of the loading of the various parcels was not, in the event, of crucial importance in view of our conclusions as to the impossibility of trying to attribute particular rats to particular parcels of cargo. However, this table indicates the extent to which parcels of SBMP belonging to particular Shippers were co-mingled with those of other Shippers in the same holds. It also shows the extent to which loading of different cargoes was taking place into different holds simultaneously.

28.

We turn now to the evidence of the presence of rats on board the vessel as it stood by the conclusion of the hearing. In so doing we note that there was an agreement between the experts that it was unlikely that any rats were present in the holds when the vessel arrived at Paranagua. We endorse that conclusion.

29.

As Counsel for Bunge emphasised, there was ample and graphic evidence (which it is unnecessary for us to recite since there was agreement between the experts on this point) that the terminals and warehouses in the Export Corridor from which the cargo was loaded at Paranagua – and the conveyor belts used to load the cargo – were heavily infested with rats. Furthermore, the rats found in the Export Corridor were of the same species as those found on board the vessel (Rattus Rattus) and the grain expert who gave evidence on behalf of the respondents, Mr Walker, accepted that some rats could and would remain on conveyors – which thus could serve as a possible conduit on to the ship.

…

33.

… By the conclusion of the evidence at the hearing it was clear that the first definite sighting of a rat – as recorded by the Master – was on 23 January 2004. …

34.

…We were inclined to agree with the submission made on behalf of Bunge that it might reasonably be inferred from the ‘fortuitous’ sighting of one rat that there were other earlier ‘missed’ occurrences of rats and that once one sighting had been made, the probability of further sightings increased dramatically because the crew would start to be more vigilant in looking for rats.

35.

Bunge went on to argue that the fact that the majority of sightings were made later on in the loading process should not be allowed to let the earlier shippers ‘off the hook’. This point is of greater relevance in the context of determining the issue as to which shipper was responsible for any particular rat sighted, but even though the standard of proof which we had to apply was only on the balance of probabilities, we considered that it would be unsafe to conclude by inference – when only relatively modest numbers of rats had been observed in total – that there must have been more. In our view it was only safe to proceed on the basis of actual sightings of rats.

36.

Although the Master did not personally see the first rat recorded as having been sighted on the evening of 23 January, he was told about it the next morning by the Chief Officer. A rat had apparently fallen on the Second Officer who was on duty at the time. We note here that the Master’s evidence was that he had been instructed by his owners that if only one rat was observed and it was caught, no mention should be made of this in the deck log. Whatever criticisms there may have been of the Master’s evidence, we were satisfied that he honestly believed that a rat had been seen by the crew in No. 2 hold late in the evening of 23 January.

…

39.

We had no reason to doubt the Master’s evidence that he was extremely concerned that many more rats had been loaded with the cargo than had been observed. However, once a decision had been taken on behalf of the vessel that the rat problem was too significant to be ‘hushed up’, it did appear that the Master took the most pessimistic view possible in recording the evidence. Thus, the vessel’s deck log for the period 23 to 26 January referred to ‘vermin loaded with cargo of soya bean meal pellets through shore loader in No. 2 hold’ and ‘nine rats were found in cargo holds amidst the …’. The reality seemed to be that no more than one or two rats had been found in each hold – though six were actually killed during the morning of 26 January when each of the holds containing SBMP was opened and inspected by the vessel’s agents and two surveyors.

40.

Following the fumigation of all the holds on 27 January, the Master informed Bunge of a further sighting of live rats on the deck on 5 February. In fact, in cross-examination the Master accepted that it was one of the members of the crew who had seen the rats on this occasion. When this was reported to Bunge they were concerned that the fumigation might not have been effective. However, the Master’s evidence in cross-examination was that his own view was that it was likely that these rats had jumped down from the structure of the ship loaders rather than come in with the cargo itself.

41.

On 8 February the vessel was inspected by pest control contractors, DACOM. Their primary concern was apparently to establish the effectiveness of the fumigation, but they also carried out an inspection of the sanitary conditions on board and reported that these were ‘very good’. No traces of rats were found in holds Nos. 1 and 2 (as well as the holds containing maize) but droppings and a carcass were found in No. 4, ‘numerous tracks on the cargo’ and a carcass in No. 6 and tracks, rodent droppings and a carcass in No. 7. They laid some rat traps which were inspected on 9 and 10 February but no further rats were found. This was despite the fact that there were significant quantities of cargo residues on the decks.

42.

The upshot of this evidence was that notwithstanding the fact that there was infestation of an alarming order ashore, the number of rats which it was possible to conclude on a balance of probabilities had gained access to the vessel was certainly less than twenty.

…

52.

… [Shippers] emphasised that no bodies of dead rats were found within the cargo at all (all the bodies found being on the surface) and none were found at Lisbon when a significant quantity of the cargo was discharged into barges for examination.

53.

It was not clear precisely what the operation carried out at Lisbon had involved. It appeared that approximately one-sixth of the cargo in each hold had been discharged by grab to barges and moved around and examined. Clearly, only the upper part of the stow was disturbed…

54.

It was also suggested on behalf of the Shippers that although Bunge sought to argue to the contrary, the Iranian receivers were clearly aware that there was a possible problem with the cargo, since they were represented by independent surveyors at Paranagua and these surveyors, Cotecna, reported that they had seen rats themselves. This information must, it was suggested, have been conveyed to the Iranian receivers and, if it had been, it was highly likely that they would have inspected the cargo carefully as it was being discharged. The absence of any complaints by the Iranian receivers was therefore a matter of some significance. There was the further possibility that any rats in the cargo would have been sieved out when the SBMP passed through a grid on its way into the animal feed plants which were apparently the end users.”

The arbitrators proceeded to consider the evidence given by experts on rat behaviour. This included such topics as the terminal velocity of a rat after leaving the final section of the loading spout, considered with a view to determining whether the rats would be likely to be killed by the descent into the hold. They discussed the native wit of rats, the importance which rats attach to hiding and their likely willingness to accept the challenge of circumventing rat guards in order to gain access to foodstuff when that was readily available ashore. The final conclusion of the arbitrators, by a majority in all but one of the references in which latter case the arbitrators were unanimous, was that the rats found on board the vessel had taken the easiest and obvious route, via the conveyor belt and the ship loader in the cargo stream, by which was meant the stream of SBMP, there being as the arbitrators also found, no rat presence noted in the maize cargo loaded in Holds Nos. 3 and 5.

The arbitrators then proceeded to address the two legal issues with which this appeal is concerned. I must again set out their discussion and conclusions:

“For each of the rats for which ingress in the SBMP cargo can be established, is it possible to prove when or by whom that rat was loaded?

67.

Given the absence of direct evidence the answer to this question was clearly ‘no’ unless it was possible to conclude that the extent of the infestation was such that on a balance of probabilities there must have been rats in every parcel of cargo. As was emphasised on behalf of the Shippers, each hold contained cargo from a number of different Shippers and as the following table reveals, there were more shippers than rats.

Hold No.	Total Rats	Shippers
1	2	3	(ADM, Rutherford c/o Granol, C&A Modas)
2	1	6	(Citrovita, Coimex, Glencore, C&A Modas, Rutherford, Orlandia)
4	3	5	(C&A Modas, MGT, Glencore, Rutherford c/o Granol, Orlandia)
6	3	5	(Glencore, MGT, ADM, Rutherford, Rutherford c/o Granol)
7	2	5	(ADM, Glencore, Rutherford, Rutherford c/o Granol, Orlandia)

68.

We have already recorded our agreement with the submission that we could not safely infer that rats must have come on board the vessel before the first sighting on 23 January. Counsel for the first four Shippers submitted that as a matter of simple timing, even if the rats came on board with the cargo, this would have meant that it was unlikely that any rats were loaded with the ADM, C&A, Citrovita or Coimex cargoes since these were loaded two days, three full days, one full day and three and a half days prior to the sighting of the first rat. We agreed that it was unsafe in the absence of direct evidence to draw inferences so that the claims against all of these Shippers must indeed fail. (Footnote: ²)

69.

However, even if we were to reach that conclusion, Counsel for Bunge submitted that the later Shippers (i.e. those whose cargo had been loaded from 22 to 24 January) – C&A, Glencore, MGT (Footnote: ³) and Rutherford – must be implicated. Bunge’s longstop position was that even if the case against all those Shippers was found to be unproven, on any view Glencore must be liable as the shipper who had completed loading immediately prior to the first sighting. (Footnote: ⁴)

70.

As the shippers understandably emphasised, even Bunge’s own expert, Dr Sheard, acknowledged that it was impossible to establish in the circumstances that a rat was shipped by each Shipper. He accepted that the distribution of any rats in the parcels of cargo could not be assumed to be even and that it was possible that whilst there might be several rats in one parcel, there would be many parcels which contained no rats.

71.

Not only did the simple arithmetic preclude a conclusion that at least one rat had entered with every parcel of cargo but there was also the possible complication that rats might have climbed from one hold to another. The reality, it was submitted on behalf of the Shippers, was that we had no evidence on which to base a proper finding on this question so that for this reason alone, Bunge’s claim against each and all of the Shippers must fail. We agreed.

B. Is a cargo loaded with a rat a dangerous cargo?

72.

As we have already outlined, Bunge’s case was that if it established on the facts that the rats boarded the vessel during cargo operations and that those rats came from the loader, the following legal consequences resulted. First, the Shippers were in breach of Article IV Rule 6 of the Hague Rules in that they had loaded a cargo which was dangerous. There was no dispute that the Hague Rules had been incorporated into the bills of lading (which were on the Congenbill 1994 Form) pursuant to the General Paramount Clause in Clause 2 of the Conditions of Carriage on the reverse side of the bills of lading.

73.

Bunge also relied upon an implied term which was pleaded as follows:

‘Further, or alternatively, it was an implied term of the contracts of carriage, by operation of common law and/or on the grounds of business efficacy, that the Respondents would not ship good of a dangerous nature, namely goods which were liable to cause damage, detention or delay to the vessel and/or to other cargo and/or to the cargo itself and/or delay in the performance of subsequent normal operations, without giving notice of such dangerous characteristics to the claimants prior to shipment.’

74.

However, in its closing written and oral submissions Bunge did not develop its arguments concerning the implied term which was, in any event, denied by the Shippers. We therefore treat Bunge as having abandoned the plea.

75.

Bunge did nevertheless rely upon the common law principle which it maintained survived the incorporation of the Hague Rules. It said that Clause 34 of the charterparty did not exclude the common law and indeed was not inconsistent with the usually implied common law terms. The clause was headed ‘Cargo Exclusions’ and contained the words:

‘… the Charterers agree that the following cargoes are excluded from carriage … All injurious inflammable or dangerous and/or injurious and/or inflammable and/or corrosive cargoes … and any other cargoes affecting immediate or long terms safety of the vessel …’

76.

Bunge submitted that a cargo which contained rats or had rats on its surface (whether those rats were alive or dead as a result of fumigation) was physically dangerous to both the vessel and/or to other cleanly loaded cargo. The danger to the vessel was suggested as the possibility of it becoming a ‘plague ship’; to other cargo there was said to be the risk of infestation. It said that the presence of even one rat was an infestation and rendered the cargo dangerous. The Joint Memorandum signed by the experts stated that ‘any level of rat infestation is potentially dangerous’ but did not elaborate further on what was meant by dangerous. Bunge’s position was, as we have already noted, that any number of rats were ‘dangerous’ and that there could not be a minimum level of dangerous infestation.

77.

In The ‘Giannis K’ [1998] 1 Lloyds Rep 337 the House of Lords saw no reason to confine the meaning of the word ‘dangerous’ to goods which were liable to cause direct physical damage to the goods. In that case (which was concerned with khapra beetle) the shipment and voyage was to countries where the imposition of quarantine and an order for dumping of the entire cargo was to be expected. Bunge argued that as the cargo on board the ‘Darya Radhe’ contained live or mummified rats (together with faeces, urine and other related by-products) which would have exposed the vessel to an arrest in Iran and may have led to the rejection of the cargo, the circumstances of the claim fell within the scope of what the House of Lords had in mind in The ‘Giannis K’. The evidence was that the cargo had not in fact been rejected although it was not clear whether that was because the cargo was clean or because of the arrangements which Bunge reached with the receivers.

78.

The Shippers said that khapra beetle had been ‘part and parcel’ of the cargo in The ‘Giannis K’ to the point where they had been inseparable. In this arbitration, however, they argued that the rats and the cargo were separate and easily separable. This submission does have a superficial attraction but does not withstand serious scrutiny in that the rats boarded as a part of and with the cargo. Nor were the rats easily separable because the evidence was that it could not be stated with any certainty at the loadport the depth of the cargo at which the rats might be found either through entering the holds at a particular time during loading or burrowing down from the surface. The rats cannot be considered as a ‘separate cargo’.

79.

The Shippers accepted, correctly in our view, for the purpose of the Hague Rules that the meaning of the word ‘dangerous’ was not confined to or had to be considered as eiusdem generis with ‘explosive’ or ‘inflammable’. However, they said that the danger had to be something intrinsic to the cargo. There is considerable force in that construction. Article IV Rule 6 in its opening words specifically refers to the nature of the shipment. SBMP by its nature is not dangerous. It can be defective and this cargo probably was defective, but that is not the same as it being dangerous in itself.

80.

It was submitted on behalf of the Shippers that if we were to conclude that the cargoes were not physically dangerous (as we have done) the claim must fail since it was not sufficient for Bunge to show that a breach of the common law obligation arose from the shipment of cargo which might possibly be subject to some legal challenge at the discharge port. That submission struck us as undeniably correct and the decision in Transoceanica v H.S. Shipton [1923] 1 KB 31 (when the court rejected an argument that a grain cargo containing stones was legally dangerous because the stones might clog up the discharging apparatus and thus cause delay) was cited as authority for this proposition.

81.

We concluded therefore that even if Bunge had succeeded in establishing that one or more rats had entered one or more of the parcels of cargo loaded at Paranagua, it would nevertheless have failed in its claim against the particular shipper(s) concerned since it was not in a position to establish a breach of contract whether by reference to the express terms of the Hague Rules or by reference to common law principles.”

The arbitrators then proceeded to make their findings on quantum, in case that should become relevant. Bunge’s claim was made under five heads:

Mr Steven Berry QC for Bunge relied upon certain of the observations made by the arbitrators when dealing with quantum as casting doubt upon their earlier conclusions on liability. I have already referred to some of these observations above when discussing the arbitrators’ finding that the cargo was discharged in Iran without incident or rejection. In relation to the argument about the diversion to Lisbon, where there was a terminal in which Bunge had an interest, the arbitrators said this, at paragraph 99 of their Reasons:

“We could not accept the submission made on behalf of the shippers that there was no likelihood of rejection of the cargo at the destination owing to the presence of rats but it struck us as a matter both of first impression and further reflection that the decision to deviate to Lisbon (with the very significant financial consequences which that ultimately involved) was impossible to justify.”

I agree with Mr Berry that this amounts to a finding that notwithstanding the steps taken there was a risk of rejection of the cargo. I return to this at paragraph 32 below. Mr Berry also not unnaturally placed reliance on the finding at paragraph 125 of the Reasons that “the consequences of the vessel arriving in Iran with live rats were of such magnitude that it was difficult to criticise a decision to re-fumigate [at Lisbon].” The arbitrators did however point out that with hindsight the risk of there still being live rats in the cargo at this stage appears to have been non-existent. This was no doubt allied to their finding that by the close of the hearing it was accepted that phosphine fumigation can be expected to be 100% effective and that it is usual practice for vessels fumigated with phosphine to sail under gas and ventilate en route, which is what Bunge had intended to do in the present case. Finally, in the same connection Mr Berry of course relied upon the arbitrators’ finding that it was reasonable for Bunge to send a representative to the discharge port.

Dangerous cargo

Mr Berry submitted that cargo is dangerous if it is liable to cause delay to the vessel and/or to the carriage of other cargo. Mr Berry suggested that this principle is to be extracted from the judgment of Atkin J in Mitchell, Cotts & Co. v Steel Brothers & Co. Ltd [1916] 2 KB 610 at page 614 where he spoke of there being at the least an obligation on a shipper not to ship goods likely to involve unusual danger or delay to the ship without communicating to the owner facts which are within his knowledge indicating that there is such a risk. There was here, submitted Mr Berry, a risk of rejection of the cargo which did cause delay. The fact that fumigation can be expected to kill all rats does not mean that there will not be an unusual delay, as here occurred. The re-fumigation, found to have been a reasonable thing to do, was on any showing an unusual delay.

Miss Sara Cockerill for her part pointed out that it is a commonplace that cargo is loaded which carries with it a risk that it will cause delay to the ship or to other cargo. Sprouting potatoes was one example which she gave, whilst pointing out that there are always awkward receivers in awkward countries. She suggested that as long ago as 1922 in Transoceanica Societa Italiana di Navigazione v H.S. Shipton & Sons [1923] 1 KB 31 McCardie J had pointed out that if the rule as to dangerous goods is extended to matters not involving physical danger, “a wide vista of responsibility is opened as against the shippers of goods”. She suggested that whatever criticism might be made of the manner in which the arbitrators expressed their reasoning, they had nonetheless come to the right conclusion.

In order to evaluate these submissions it is necessary to return to first principles. Fortunately those principles have in large part recently been explored and restated by Longmore J, the Court of Appeal and the House of Lords in Effort Shipping Co. Ltd v Linden Management S.A., The “Giannis NK” [1994] 2 Lloyds Rep 171, [1996] 1 Lloyds Rep 577 and [1998] AC 605.

In The “Giannis NK” at first instance Longmore J drew attention to the danger of confusing two distinct but allied principles in relation to the shipment of goods which cause loss to a shipowner. At page 179 he said this:

“The common law of England and Wales has for some time recognized two similar but distinct principles in relation to the shipment of goods which cause loss to a shipowner. There is first the principle that a shipper undertakes not to ship goods which are liable to cause damage to the vessel or other cargo shipped thereon without giving notice to the shipowner of the character of the goods. This proposition was laid down in, for example, Brass v. Maitland, (1856) 6 El. & Bl. 470, where the shipper was held potentially liable for shipping casks of bleaching powder which, unknown to the shipowner, contained lime chloride which damaged the casks and leaked out causing damage to other cargo on board.

There is also a second principle that the shipper undertakes not to ship goods which are liable to cause delay to the vessel. Examples of such goods are contraband cargo or cargo for which the cargo-owner requires a licence for import which he does not, in the event, obtain. An example of this second principle is Mitchell v. Steel, [1916] 2 KB 610, in which charterers had loaded a cargo of rice for Alexandria but then asked the shipowners to discharge the cargo at Piraeus instead. The shipowners agreed to this course but the cargo could not be discharged at Piraeus without the express permission of the British Government who controlled the port in the First World War. Permission was not granted and the vessel had to discharge at Alexandria after all. The charterers were held liable for the resulting delay by Mr Justice Atkin, who described the case as being analogous to a case of dangerous goods.

The second principle has been treated by authors and editors of books on carriage by sea as part of the law concerning dangerous goods and it seems that both principles may indeed have a common origin in a proposition set out in the fifth edition of Abott on Shipping (1825) at p. 270:

The general duties of the merchant … are comprised in a very narrow compass. The hirer of anything must use it in a lawful manner and according to the principle for which it is let. The merchant must lade no prohibited or uncustomed goods by which the ship may be subjected to detention or forfeiture.

See, also, the current 14^th edition, published 1901, p. 643.

This passage was cited by both the majority judgment in Brass v. Maitland, (1856) 6 El. & Bl. 470 at p. 484 and the minority judgment given by Mr Justice Crompton, at p. 492, which Mr Justice Atkin preferred to follow in Mitchell v. Steel. After successful Counsel for the shipowners in that case himself became the Commercial Judge, he also referred obiter to the above statement of Lord Tenterden as constituting the true basis of the doctrine which ‘is apt to be a little obscured if one thinks only of dangerous goods’, see Rederii Aktiebologet Transatlantic v. Board of Trade, (1924) 20 Ll.L.Rep. 241 at p. 243, col. 1; (1924) 30 Com. Cas. 117 at p. 128.

This brief account of the development of English law helps to put Mr Matthews’ first two submissions into context and shows that there is a danger of confusing what are two distinct but allied principles.”

Longmore J concluded that the words “goods of a dangerous nature” in the Hague Rules are probably restricted to goods which are physically dangerous, i.e. liable to cause physical damage to some object other than themselves. In that case the Khapra beetle infested cargo, ground-nut extraction meal pellets, did not cause physical damage to the vessel. However it caused damage to the other cargo, wheat, because the shipment and voyage was to countries where the imposition of a quarantine and an order for the dumping of the entire cargo was to be expected. As a result of the exercise by the authorities of their statutory powers controlling importation, the other cargo had eventually to be dumped at sea. As Hirst LJ put it in the Court of Appeal, adopting the argument of Mr Alistair Schaff QC, the whole cargo, including the wheat, was blighted. From the moment the ground-nut cargo was loaded the wheat cargo was subjected to the physical peril of being dumped. The decision of the House of Lords on this point is summarised by the learned editors of Carver on Bills of Lading, 2^nd Edition, at paragraph 9-280 as follows:

“It was held by the House of Lords that the word [dangerous] does not require that the goods cause or become capable of causing direct physical damage to persons, the ship or even to other cargo, and that it applies also to cargo liable to give rise to (physical) loss of other cargo on the same vessel by creating a situation (in the case in question, infestation) in which it was required by public authorities that all the cargo be dumped at sea. It appears that the words refer to cargo which directly or indirectly causes some sort of physical damage to life, the ship or other cargo, or raises a threat of it (for instance, in the case of a cargo of explosives), leading to delay and/or other expense for the carrier. Goods that merely cause delay to the carrier, for example because their import is prohibited, would probably not be within these words, though as stated in the previous paragraph they might rank as dangerous goods at common law.”

It is plain therefore that goods may be dangerous for the purposes of the Hague Rules if they have the capacity to cause physical damage in either a direct or an indirect manner. However as the learned editors of Carver point out, goods that merely cause delay to the carrier are probably not to be regarded as “dangerous” in the sense in which that word is used in the Hague Rules. Since all three courts in which The “Giannis NK” was considered concluded that the ground-nut cargo was the indirect cause of physical damage to the wheat cargo, it was unnecessary to decide whether goods which merely cause delay to the carrier fall within the Hague Rules definition of dangerous goods. Longmore J pointed out, at page 180 LH, that in the light of his earlier exposition, the owners would not in that case have been without a remedy even had he been unable to conclude that the cargo was physically dangerous, in the sense that it had the capacity to cause physical damage in either a direct or an indirect manner. However nothing said in any of the judgments or speeches in The “Giannis NK” gives any encouragement to the view that the word “dangerous” can in the Hague Rules carry a wider meaning than that discussed above. The learned editors of Scrutton on Charterparties and Bills of Lading, 21^st Edition, at page 409, are at one with Carver in thinking that “dangerous” in Article IV Rule 6 “probably means physically dangerous, and does not extend to those cases where the ship suffers loss owing to legal obstacles to the carriage or discharge of goods”. I would however just observe that it is in my view most unlikely that the word “dangerous” can be intended when used in Article IV Rule 6 of the Hague Rules to bear a meaning going beyond physical danger. The owner has the right under the Rule at any time before discharge and without incurring a liability to pay compensation to land “dangerous” cargo at any place or to destroy it or to render it innocuous. Quite apart from the obvious pointer given by the expression “render it innocuous” it would be very surprising if the owner had the right without incurring any liability whatsoever either to land at a non-contractual destination or even to destroy cargo which posed no physical threat to either ship or other cargo carried.

Irrespective of the Hague Rules, there is of course at common law in a contract of carriage an implied term to the effect that the shipper will not ship dangerous goods without notice to the carrier. That is the first principle mentioned by Longmore J in the passage cited above. It is in my view clear that the meaning of the word “dangerous” in this term implied by the common law is the same as that which has been given to the same word in Article IV Rule 6 of the Hague Rules. The obligation is absolute, as is the liability of the shipper under Article IV Rule 6 – see The Giannis NK in the House of Lords per Lord Lloyd of Berwick at page 619 and Lord Steyn at page 624.

I turn then to the second principle mentioned by Longmore J. This principle is shortly summarised in Scrutton at Article 53, page 100 as being a particular incident of Longmore J’s first principle:

“Goods may be dangerous within this principle [Longmore J’s first principle] if owing to legal obstacles as to their carriage or discharge they may involve detention of the ship.”

The authority for this proposition is given as Mitchell Cotts v. Steel and The Giannis NK. In Mitchell Cotts v. Steel Atkin J expressed the principle in this way, at 614:

“I think there is no question that a shipment of goods upon an illegal voyage – i.e., upon a voyage that cannot be performed without the violation of the law of the land of the place to which the goods are to be carried – a shipment of goods which might involve the ship in danger of forfeiture or delay – is precisely analogous to the shipment of a dangerous cargo which might cause the destruction of the ship. I do not think there is any distinction between the two cases.”

The formulation in Scrutton of this principle, or sub-principle, is in my view instructive. The learned editors refer to legal obstacles. It was the local law at the port of discharge with which Atkin J was concerned in Mitchell Cotts v. Steel and to which he referred in the passage cited above. Charterers there loaded a cargo of rice for carriage to Alexandria, which at the time, 1915, was under British control. On passage from the load port it was agreed to vary the charterparty by changing the contractual destination to Piraeus. When negotiating with the owners for this variation the charterers were aware of the fact that they could not send the ship to Piraeus with the cargo of rice on board without the consent of the British Government. The shipowners did not know this and could not reasonably have known it and the charterers did not tell them. Whilst the charterers were endeavouring to obtain the relevant consent the vessel was detained at Port Said. Consent was not forthcoming and the cargo had in the event to be discharged at Alexandria. The owners claimed for detention at Port Said. The claim succeeded in arbitration and Atkin J upheld the award on an appeal by way of case stated. As Longmore J observed, Atkin J treated the case as analogous to one of shipping dangerous goods. Longmore J for his part did not discuss the ambit of this second or sub-principle, although the examples he gave of goods liable to cause delay to the vessel were contraband cargo or cargo for which the cargo owner requires a licence for import which he does not, in the event, obtain. Those are both examples of “legal obstacles” as to the carriage or discharge of the goods. Longmore J had already pointed out, at page 173 of his judgment, that both the USA and many Caribbean countries, including the Dominican Republic, the countries there concerned, had for many years exercised their statutory importation powers to exclude vessels and cargo infested with Khapra beetle. Thus when he later observed that in that case the owners were not without remedy even if the cargo was not physically dangerous, he was adverting simply to the fact that the case was in any event on all fours with Mitchell Cotts v. Steel in that it involved a violation of the law of the land of the place to which the goods were to be carried.

In The “Giannis NK” the House of Lords agreed with the courts below that the Khapra beetle infested cargo was “physically dangerous” because it was liable to give rise to the loss by dumping at sea of other cargo loaded on the same vessel. It posed, albeit indirectly, a physical danger to the other cargo. In those circumstances, as Lord Lloyd observed at page 613:

“… It is unnecessary to consider a further argument that goods may be of a dangerous nature even though they do not present any physical danger to ship or cargo, but are ‘legally’ dangerous in the sense that they are liable to cause delay to ship and cargo through the operation of some local law.”

The Court of Appeal in that case, having agreed with Longmore J that the relevant cargo was “physically dangerous” to the other cargo, went on to say that it would in any event uphold Longmore J’s judgment by reference to the principle stated in Mitchell Cotts v. Steel “seeing that, on any view, the shipment of the infested cargo was likely to involve detention and delay of the vessel, as in fact occurred”. See per Hirst LJ at page 588, with whom both Morritt and Ward LJJ agreed.

Mr Berry sought to extract from the judgment of Atkin J a principle much broader than that mentioned, although not discussed, by Lord Lloyd in The Giannis NK. He suggested that the principle enunciated by Atkin J was not dependent upon the existence of any legal obstacle such as one imposed by the local law at the port of destination. He founded in particular on the following passage in Atkin J’s judgment, at page 614:

“Whatever may be the full extent of the shipper’s obligations, it appears to me that it amounts at least to this, that he undertakes that he will not ship goods likely to involve unusual danger or delay to the ship without communicating to the owner facts which are within his knowledge indicating that there is such a risk, if the owner does not and could not reasonably know those facts. I think that is placing the obligation of the shipper within very moderate limits, and it may be considerably wider.”

In my judgment Mr Berry is misreading this passage and ignoring its context. Atkin J had in an earlier passage expressed his preference for the dissenting judgment of Crompton J in Brass v. Maitland, in which he had rejected the view of the majority that the duty at common law not to ship dangerous goods is an absolute duty, expressing his own view that the duty could only be one to take due and proper care and diligence not to ship dangerous goods without notice to the carrier. Then in the passage I have cited at paragraph 27 above Atkin J expressed the view that shipment of a cargo upon what he termed “an illegal voyage” is precisely analogous to the shipment of a dangerous cargo which might cause the destruction of the ship. The passage upon which Mr Berry relies immediately follows. It is plain in my judgment that Atkin J was not there intending to state a principle which was broader than that which he had stated in the previous paragraph – indeed in the previous sentence but one. That was a principle which involved illegality in performance. The later passage is intended to convey that whether the underlying obligation should be regarded as absolute or more limited, at the very least a shipper is liable in circumstances where he knows of the facts which give rise to the illegality and fails to inform the carrier who neither knows nor could reasonably know the facts out of which the illegality arises. That is why Atkin J concluded the passage in question with his remark “I think that is placing the obligation of the shipper within very moderate limits, and it may be considerably wider,” by which no doubt he meant to suggest that the obligation might in fact be absolute rather than a duty to take care. Furthermore, as Miss Cockerill pointed out, in The Lisa, [1921] P 38, Sir Henry Duke P held reliance upon the judgment of Atkin J to be misplaced in a case where the risk of seizure of the cargo arose not from any municipal law or regulation in force at the two ports concerned, Kirkwall and Narvik, but rather from the possibility of the British Government seizing the ship as prize in exercise of belligerent rights.

In Transoceanica v. Shipton McCardie J regarded the decision of Atkin J in Mitchell Cotts as “undoubtedly enlarg[ing] the duty of a shipper” because the cargo of rice was not a dangerous cargo in itself. He considered that Atkin J had “gone one step beyond the dangerous goods principle”. He pointed out that “if the rule as to dangerous goods is extended to matters which do not involve danger, a very wide field is opened for discussion”. I do not need to discuss whether Atkin J’s decision in fact enlarged the duty beyond that which had been recognised in relation to the shipment of goods involving some illegality in performance. If regarded however as analogous to the shipment of dangerous goods, the principle espoused by Atkin J undoubtedly goes further than does the common law approach to physically dangerous goods, and if not confined to cases of “legal obstacle” would as McCardie J observed open a wide vista of responsibility. In the Shipton case a cargo of barley contained sand and stones which choked the suction pump used on discharge and thereby caused delay. McCardie J rejected an argument that the shippers were liable by analogy with the liability for the shipment of dangerous goods.

Accordingly I reject Mr Berry’s submission that the principle stated by Atkin J in Mitchell Cotts operates independently of legal obstacle, and I likewise reject his associated submission that shippers are liable if they load a cargo which is at risk of rejection which in turn causes cost and delay, in the shape here of the time spent fumigating the cargo and otherwise dealing with the cargo in a manner designed to reduce or eliminate the risk. In truth all or most cargo is at risk of rejection on discharge, whether justifiably or not, and the allocation of the risk of delay arising therefrom is dealt with in contracts of carriage quite independently of the regime as to dangerous cargo. The principle in Mitchell Cotts is concerned with the violation of or non-compliance with some municipal law which is of direct relevance to the carriage or discharge of the specific cargo in question.

Mr Berry submitted that the rats in the SBMP cargo presented an obvious physical danger to the maize cargo in that they were likely to eat it, excrete upon it and spread disease to it. He also submitted that the rats presented obvious legal danger in the shape of the risk, particularly in a voyage to Iran, of the arrest of the ship and condemnation of all cargo on board, reminding me of what the arbitrators had said about the consequences of the vessel arriving in Iran with live rats. The difficulty about these submissions is that they are precluded by the facts found by the arbitrators. Furthermore, Bunge did not before the arbitrators rely upon any principle of Iranian or indeed any other system of law, and there are accordingly no findings on that topic. The arbitrators specifically found that no rat presence was noted in the maize cargo in Holds Nos. 3 and 5. There is no express finding that the maize cargo was even at risk of rejection. Even if the arbitrators’ remarks at paragraphs 99 and 125 of the Reasons are to be taken as referring to all of the cargo, including the maize, they must be read in the context of the entirety of the arbitrators’ findings. The arbitrators found that fumigation was in any event routine. It could be expected to be 100% effective. “Mummified rats” were no more than a cosmetic problem. Bunge was no doubt genuinely concerned and could not be criticised for re-fumigating, but looked at objectively with the benefit of hindsight its concern was in fact groundless – the risk of a rat surviving the first fumigation was “non-existent”. At paragraph 99 of their Reasons the arbitrators did say, as I have mentioned above, that they could not accept the submission of shippers, made in relation to the deviation to Lisbon, that there was no likelihood of rejection of the cargo at the destination owing to the presence of rats. However the Reasons must be read as a whole in a fair, reasonable and generous manner. As Bingham J said in Zermalt Holdings v. Nu-Life Upholstery Repairs, [1985] EGLR 14, the “courts do not approach [arbitration awards] with a meticulous legal eye endeavouring to pick holes, inconsistencies and faults … and with the objective of upsetting or frustrating the process of arbitration. Far from it. The approach is to read an arbitration award in a reasonable and commercial way, expecting, as is usually the case, that there will be no substantial fault that can be found with it.” Thus I do not consider it fair or reasonable to read this observation of the arbitrators as intended to be a considered conclusion which would be at odds with the broad thrust of their detailed findings, which I have summarised above. I suspect that they meant to say no more than that whilst the fumigation could be expected to be 100% effective, some residual risk could not entirely be discounted. I do not think that by not accepting the submission of no likelihood of rejection they intended to convey a finding that rejection was likely, particularly since in the same sentence they found the decision to deviate to Lisbon impossible to justify. Leaving aside the finding that the decision to deviate to Lisbon was in any event dictated by concerns about the bunkers rather than concerns about the cargo, it is in my view plain that the arbitrators considered that whilst Bunge was not to be criticised for re-fumigating, they nonetheless regarded it as an over-reaction. The key to understanding much of what the arbitrators said is I suspect to bear in mind that Bunge in relation to this venture wore two hats. For the purposes of the contracts with the shippers, it was the carrier of the goods. However Bunge was also, as expressly found, the ultimate purchaser of the cargoes being shipped and, as I infer, was the seller of the same to the Iranian interests. Much of what Bunge did was no doubt reasonable in its own interests as purchaser and seller of the cargo but was not necessarily the reaction to be expected of a reasonable carrier with no interest in the cargo itself.

For all these reasons as it seems to me Bunge simply lacked the factual findings on the basis of which it could hope to bring home liability to the shippers on the footing that the cargo shipped was dangerous. The arbitrators positively found that the cargo of SBMP did not pose a physical danger to the maize cargo. It plainly posed no threat of damage to the ship itself. The arbitrators made no finding that imposition of quarantine or dumping of the entire cargo was to be expected. They made no finding that the cargo was likely to involve the ship in unusual danger or delay, still less that such danger or delay would arise in consequence of the existence of some legal impediment to the carriage or discharge of the cargo. In my judgment in the light of their factual findings the arbitrators correctly applied the law. It may be that when they explained their reasoning they elided some of the concepts which I have discussed above. It is possible that in paragraph 79 the point that they were seeking to make is that the cargo, even with rats admixed, did not have some intrinsic property such as being inflammable or explosive or corrosive such as to render it likely to cause direct damage to the ship or to other cargo. That of course would not be conclusive of its ability to cause physical damage in an indirect fashion, or by “blighting” the other cargo with the risk of loss through dumping. But there are findings which preclude a conclusion that the cargo had this quality. As to the Mitchell Cotts principle, the arbitrators were quite right to say at paragraph 80 of their Reasons that it is insufficient to demonstrate a breach to prove simply shipment of a cargo which might possibly be subject to some legal challenge at the discharge port. It is not unusual for a cargo to be the possible subject of rejection or to give rise to a risk of arrest of the vessel by receivers. No attempt had been made before the arbitrators to demonstrate that shipment of the cargo in the condition in which it was shipped carried with it the risk of delay and expense in consequence of the operation of some specific municipal law or regulation.

Mr Berry did somewhat faintly suggest that the matter might be remitted to the arbitrators for further findings but this would be futile in the absence of further evidence. It would also be contrary to principle since the case was not argued before the arbitrators by reference to any specific local law. In any event the arbitrators’ findings in my view preclude the possibility of success of any argument based on local law, since the thrust of their findings is that fumigation was routine and intended and could be expected to eliminate all but cosmetic problems, which it did.

For all these reasons I conclude that the arbitrators made no error of law but on the contrary came to the correct conclusion that Bunge could not establish a breach of contract, or liability under the Hague Rules, even assuming that it could show that one or more of the shippers was responsible for the introduction of one or more rats. The second question of law does not therefore arise. I will however go on briefly to consider that question, which is essentially whether the arbitrators erred in their approach to the discharge of the burden of proof. I would start by observing that it would be surprising if experienced arbitrators fell into this error.

Discharge of the burden of proof

In my judgment the arbitrators approached their task in the orthodox and correct manner. Bunge’s dilemma was apparent from the manner in which it pleaded its case against the shippers, as the arbitrators record at paragraph 4 of their Reasons. Bunge’s case was:

“Many live rats were loaded by each, alternatively several, alternatively one or more of the Respondent Shippers, into each of Holds 1, 2, 4, 6 and 7 with their respective cargoes.”

It is also apparent, and the arbitrators effectively find at paragraph 11 of their Reasons, that for whatever reason Bunge’s case as to the number of rats involved, at any rate as it was advanced before the oral hearing, was exaggerated. Bunge’s case at the hearing was that each and every one of the SBMP parcels was infested – Reasons paragraph 20. The evidence having demonstrated the presence of only between 14 and 20 rats, (Footnote: ⁵) that proposition became the more implausible.

At paragraph 20 of their Reasons the arbitrators formulated the issue which they had to decide as whether, if the rats had come on board the vessel with the cargo, there were rats present in every shipper’s cargo, or only some of them, and if so which? That was in my view precisely the issue which the arbitrators had to decide. Mr Berry submitted that the question which the arbitrators should have asked was “on a balance of probability, did the shipper in question load a rat?” Bunge, he said, needed only to prove on the balance of probabilities that each shipper probably loaded one rat. There is I think no difference in substance between the issue posed by the arbitrators at paragraph 20 and the question which Mr Berry submits they should have asked themselves.

Mr Berry submitted that Bunge’s point both at the arbitration and now on appeal is simple. Where 14-20 rats are loaded in cargo by nine shippers, it can and must be concluded that every shipper probably loaded at least one rat. In my judgment that is a wholly unsustainable proposition simply as stated, and it becomes even more unsustainable in the light of the facts found by the arbitrators. The only fact which supports Mr Berry’s conclusion is that 14 is a greater number than 9. He could not have made the same submission had there been 14 rats and 13 shippers.

The arbitrators were “inclined to agree with the submission that it might reasonably be inferred from the ‘fortuitous’ sighting of one rat [the first] that there were other earlier ‘missed’ occurrences” – see paragraph 34. That notwithstanding, since “only relatively modest numbers of rats had been observed in total” the arbitrators concluded that it was unsafe, in the legal context of application of the standard of proof, to infer that there must have been more rats. They concluded that it was only safe to proceed on the basis of actual sightings of rats – paragraph 35. In my judgment there is no inconsistency between these two passages in their Reasons and the arbitrators’ approach is, I consider, suitably cautious. The adage, if such it be, that where there is one rat there is more than one rat is not a sound basis upon which to found legal liability for multiple rats, and I do not think that the arbitrators were saying any more than that. The arbitrators were in my judgment perfectly entitled to conclude that they could not safely infer that rats must have come on board the vessel substantially before the first sighting late in the evening on 23 January. Indeed, in my judgment the arbitrators could legitimately be criticised had they attributed responsibility to the five shippers who had completed loading their parcels at least 24 hours and in some cases much longer before the first sighting of a rat. From this it follows that Mr Berry’s proposition resolves into “where 14-20 rats are loaded in cargo shipped by four shippers, it can and must be concluded that every shipper probably loaded at least one rat”. Although this is superficially a more attractive proposition than the first, logic still does not compel assent to it.

At paragraph 67 of their Reasons the arbitrators address the question whether the extent of the infestation was such that on a balance of probabilities there must have been rats in every parcel of cargo. That was of course the question which Bunge had required them to address by submitting that each and every one of the parcels was infested. In their table at paragraph 67 of their Reasons, as I understand it, they regarded each shipper contributing cargo to a given hold as contributing one parcel of cargo to that hold. It was on that basis, comparing I think the number of parcels loaded into each hold with the number of rats sighted in that hold, that they concluded that there were more shippers than rats. They might equally have said that there were more parcels than rats but the point is clear. Had they asked themselves whether on a balance of probabilities there must have been rats in each and every parcel of cargo which corresponded to the division of the cargo into the 30 parcels covered by the 30 bills of lading, they would obviously have reached the same conclusion. It is true that the arbitrators prefaced that section of the Reasons which began at paragraph 67 with the sub-heading question “for each of the rats for which ingress in the SBMP cargo can be established, is it possible to prove when or by whom that rat was loaded?” I would agree that the answer to that question would not in itself be determinative of the question whether liability could be established against any one shipper. Not unnaturally the arbitrators answered this question “no”. Furthermore the answer to this question would not have been “yes” even had the extent of the infestation been such that on a balance of probabilities there must have been rats in every parcel of cargo. Yet had the arbitrators concluded that on a balance of probabilities there must have been rats in every parcel of cargo, plainly they would have regarded liability as potentially established against every shipper, subject of course to the issue of dangerousness. What this demonstrates to my mind is that the question posed in the sub-heading immediately before paragraph 67 is not the only question which the arbitrators addressed. I do not think that they lost sight of the question for decision, which as they correctly posed it at paragraph 20 was whether on a balance of probabilities there were rats present in every shipper’s cargo, or only some of them, and if so which?

No doubt there might be cases of this sort where statistical evidence could be deployed in an effort to prove what on the balance of probabilities must have occurred. Here I suspect the number of rats involved is simply too small to permit any valid statistical approach. Furthermore, given that the arbitrators regarded it as unsafe to infer or to assume that rats were introduced substantially before the first sighting, and given that certain shippers completed loading long before others, and in some cases before others had even started loading, any attempt to show an even distribution of the introduction of rats over time could not give rise to an inference that all shippers were responsible for the introduction of at least one rat. As I understand it no attempt was made at the hearing to pursue some sort of statistical approach and this is unsurprising. However, as is reflected in paragraph 70 of the Reasons, Bunge’s expert witness Dr Sheard, who was apparently called to give evidence on technical aspects of loading, did offer some evidence in this area, and it was not suggested before me that he was unqualified so to do. As the arbitrators found at paragraph 70, Dr Sheard acknowledged that it was impossible to establish in the circumstances that a rat was shipped by each shipper. He accepted that “the distribution of any rats in the parcels of cargo could not be assumed to be even and that it was possible that whilst there might be several rats in one parcel, there would be many parcels which contained no rats”.

I do not agree with Mr Berry that the arbitrators approached the matter on the footing that Bunge had to show that there was no possibility that the rats were distributed unequally throughout the cargoes. Plainly they did not. Subject to dealing with the timing point, Bunge could as it seems to me have established liability against some of the shippers had they been able to demonstrate that on a balance of probabilities the distribution of the introduction of rats should be regarded as even, but they did not so demonstrate. The evidence of Dr Sheard was to the contrary effect.

Finally Mr Berry did not I think formally abandon the argument that at least Glencore should be regarded as probably having loaded a rat, being the shipper who had completed loading immediately prior to the first sighting. Mr Berry did not press this point in oral argument. It is I think unavailable to him in the light of the arbitrators’ clarification that they made no finding as to whether any particular party was the shipper whose cargo was last loaded before the first rat was sighted. The arbitrators did not suggest that they regarded the timings in Appendix A, to which they refer at paragraph 27 of their Reasons, as necessarily sufficiently accurate to establish which shipper was last loading before the first sighting, and Appendix A collates evidence from three sources, not always consistent. As it happens, Appendix A seems to suggest that Rutherford may have been loading in Hold No. 2 until 25 minutes later into the evening on 23 January than were Glencore – 2115hrs as opposed to 2050hrs, but equally the finding as to the first sighting is only that it was “late in the evening”. Had this been a live issue it might have been necessary to explore in a little more depth whether the arbitrators’ reluctance to infer that rats must have been loaded substantially prior to the first sighting necessarily leads to the conclusion that on the balance of probabilities the rat first sighted must have been introduced by the shipper or shippers who was or were loading immediately before that first sighting. As it is the point is not determinative and I need say no more about it.

For all these reasons I dismiss Bunge’s appeal against the eight awards.