In this action, the court is asked to adjudicate on claims for condensate damage to nine consignments of washed Colombian green coffee beans transported by the defendant container line (“the carrier”) in a total of twenty dry and unventilated 20 ft containers each loaded with 275 hessian bags of 70 kg from Buenaventura in Colombia to destinations in North Germany. The loading dates ranged from 1 January 2012 to 6 April 2012 (though mostly in the second half of March 2012) with delivery to the consignees between February and May 2012. In each case, the consignments were off-loaded at Balboa in Panama and transhipped into a vessel bound for North West Europe. In one case, the consignment was on-shipped directly from Balboa to Bremerhaven. In another case, on-shipment was to Hamburg followed by road-transport to, and delivery in, Bremen. The other seven consignments were on-shipped to Rotterdam, a large hub for container traffic, and off-loaded there before on-carriage by ship to Hamburg, where they were in one case delivered to the consignee and in six carried by road to, and delivered to the consignees in, Bremen. The first leg of the sea-carriage, from Buenaventura to Balboa, was on deck; the on-carriage to Europe under deck; and the further carriage from Rotterdam to Hamburg on deck. The off-loaded periods in Rotterdam and Hamburg varied considerably, for example in Rotterdam from less than one day up to fourteen days.

With one exception the claimants are the consignees under the bills of lading issued in respect of the consignments (Footnote: ¹) . The sixth claimant is the agent of their cargo underwriters, who have indemnified them against their loss; whether or not their presence in the action as co-claimants is either necessary or appropriate, it has not been challenged.

On outturn each of the consignments and all of the containers save two were found to have suffered some degree of damage from condensation. In bare essentials, moisture in warm air rising from the stow had condensed on contact with the cold roof of the container (Footnote: ²) , fallen on to bags at the top of the cargo, and also run down the sides of the container wetting bags on the outside of the stow.

The bills of lading recorded in the usual way receipt of the consignments in apparent good order and condition. There is no reason to believe, nor is it alleged by the carrier, that the beans were unusual or atypical as to their moisture content. Indeed, outturn surveys (which took place on all but one of the consignments) recorded readings which were average or better, around 11% and 12% moisture content. Nor was it suggested that they were unusual in any other respect.

The carriage was effected on LCL/FCL terms, which means that the containers are provided and filled or “stuffed” with the bags by the carrier but “unstuffed” or “stripped” by the consignee after arrival at their destination. Before stuffing, the bare corrugated steel of the container was lined by the stevedores with Kraft paper. The primary question in the present case, in essential brevity, is whether or to what extent that was properly effected and adequate to meet the threat of condensation, and, if not, whether the carrier was liable for any consequent damage (Footnote: ³) .

The damage was relatively minor in both extent and financial impact. While the proportion of bags affected in each container varied from 0% (in two cases) to as high as 56.7%, the damage concerned only a very limited part of their contents. Some of those bags, consigned to the first claimant, were sold as “damaged” at a discount to the market price; some, consigned to the fifth claimants, were accepted by them for their own use against payment of 10% of their CIF value by cargo insurers (represented by their underwriting agents, the sixth claimants). A significant part of the claims was not for the damaged beans themselves but for ancillary costs of inspection and sorting of bags containing damaged beans or showing external signs of wetting. The total value of the claims, identified separately in the Schedule to the Claim Form, was advanced as US $ 87,195.49 (compromised during closing submissions at a global figure of US $ 62,500), less than US $ 5,000 per container. Ultrasensitive antennae are not required to detect that this litigation, effectively between cargo underwriters and the carrier’s P & I club, is concerned at least as much with the wider implications of its result as with any monetary recovery, and counsel gave me to understand as much. Consonant with this, the proceedings were prepared and fought as if the claims were ten times higher in value. In particular, the parties called (between them) three experts of high calibre. Also in line with the litigation’s wider overtone, neither the arguments nor the expert evidence focussed to any serious extent on individual containers other than as exemplars, and the contents of this judgment are conditioned accordingly.

Under Condition 10 of the bills of lading the carrier undertook responsibility for the whole of the intermodal transport from the port of loading (viz. Buenaventura) to the port of discharge. Condition 2, a standard Clause Paramount, made the carriage subject to the Hague Rules as regards the carriage by sea, defined by Article I (e) of those Rules as “the period from the time when the goods are loaded on to the time they are discharged from the ship”.

As a preliminary point counsel for the carrier submitted that the stuffing of the containers by its stevedore agents occurred before the loading. Accordingly, he contended, the Hague Rules did not apply, leaving the carrier free (as recognized by Article VII) to invoke two special conditions, Conditions 11 and 12, which were said to reduce the scope of its obligation from that which would otherwise be prescribed by Article III (2). This argument fails in my judgment at a number of points.

Where cargo is loaded into a carrier’s containers which are subsequently loaded on the vessel, it is in my view unrealistic to treat this as anything other than a single loading process, even if there is inevitably some interval between the two. Even if that were not so, the parties are free to agree on what for the purpose of Article I (e) constitutes loading: see Pyrene Co. LD. v Scindia Navigation [1954] 2 WLR 1005, approved in G.H. Renton v. Palmyra Trading [1957] AC 149. Where, as here, the obligation to stuff its own containers is assumed by the carrier, I would have little difficulty in interpreting the contract of carriage as including that as part of the loading.

It is not suggested that the stuffing, i.e. the actual loading into the containers, was itself defective (Footnote: ⁴) or caused any loss (for example if – as in Pyrene - the cargo had been mishandled and damaged during the operation). Properly analysed, the complaint is rather that the bags of coffee beans were carried in the immediate physical environment of a steel container carelessly or inadequately prepared to protect the cargo against damage from condensate. The breach alleged is thus of a duty at the heart of the carriage.

If one had been concerned with an act or omission preceding the inception of the Hague regime, the carrier’s responsibilities would be either those of a common bailee, requiring it in case of loss or damage to rely on an excepted peril or disprove negligence, or a bailment on terms, in which case I would have little difficulty in holding that the obligation to stuff assumed by the carrier was to be implemented carefully and properly. In both cases the result would be similar to that under the Hague Rules and though it would yield to contrary agreement, neither of the two contractual Conditions on which the carrier seeks to rely would in my view improve its position compared with that regime.

Condition 11 provides (so far as relied upon by the carrier):

Since, as I have indicated, the bagged beans are not alleged to have been unusual, in particular as regards moisture content, the question which would arise under the first of these two paragraphs would be in effect no different from that to be addressed under Article III (2) and/or Article IV (2) (m) as judicially interpreted, which I consider later in this judgment.

The second paragraph addresses a case where the containers are stuffed by the shipper etc. (Footnote: ⁶) , which is not this case. Indeed, one may derive from it e contrario that where the stuffing is effected by or on behalf of the carrier or its agents it must be done safely and properly.

Condition 12 provides (so far as relied upon by the carrier):

Article II of the Hague Rules provides:

The word “carefully” in Article III (2) has, so far as I am aware, received no significant judicial elaboration - certainly, none was drawn to my attention – and probably requires none. “Properly” has however been construed, on the basis of its natural and ordinary meaning, as “in accordance with a sound system” ( Albacora SRL v. Westcott & Laurance Line Ltd [1966] 2 Lloyds Rep. 58 following G .H.Renton v Palmyra Trading Corporation, [1957] A.C. 149 at 166 per Lord Kilmuir L.C.). The soundness of the system is to be judged without reference to any “weaknesses and idiosyncrasies of a particular cargo” ( Albacora at 62 per Lord Pearce), since these are not matters of which a carrier ought normally to be aware. In the present case, however, the cargo was entirely typical, and the fact that the carrier dressed the containers with Kraft paper reflected a (well-documented (Footnote: ⁷) ) industry awareness of condensate damage as a general and significant risk in unventilated containerised carriage of coffee beans. The question is whether the precautions taken against such damage were both appropriate in conception ( proper ) and implementation ( careful ). Though judges have tended to address “carefully” and “ properly” as separate requirements, and that is generally a helpful analytic approach, they can also be viewed as complementary components or aspects of a single contractual duty of care, so that it may not always be necessary – or sometimes even possible – to assign a breach of Article III (2) to one or other of them in isolation. This can be of particular relevance where the onus of proof comes into play.

As regards Article IV (2), the authorities on the meaning of inherent vice classically begin with The Barcore [1896] P. 294 at 297:

That leads to the question of onus of proof. Counsel for the claimants submitted that the Hague Rules should be understood as continuing the position under the pre-existing law in which the ship was treated as a common carrier liable for any damage to the goods which it could not prove to have been caused by inherent vice or any other cause not involving its negligence. That is in my view not an appropriate way to treat such an internationally adopted code. The correct approach is rather that the forensic tool of res ipsa loquitur is as available here as in other areas of litigation where knowledge of the underlying circumstances is by their nature confined effectively to the defendant. If goods, acknowledged to have been received in apparent good order and condition, are delivered in a damaged state, that can without more be sufficient to justify the court in inferring a breach of Article III (2) and thus require evidence from the carrier to negate such a breach: see e.g. per Lord Pearson in Albacora at 63. Whether this is achieved by invocation of a specific sub-rule, such as proof of inherent vice, or the general reserve catch-all of sub-rule (q), or direct refutation of Article III (2) would appear unimportant. I need not therefore pause to consider whether inherent vice is the appropriate concept when the damage is caused by water droplets generated by the interaction of moist air with the cold ceiling of a container. In any event, however counterintuitive that might appear, the claimants have not sought to argue the contrary (Footnote: ⁸) .

As pleaded, the claimants’ case is advanced primarily as one of res ipsa loquitur (though followed by further or alternative positive allegations relating to the use and deployment of the Kraft paper (Footnote: ⁹) ). Counsel for the carrier properly accepted that the undisputed damage on outturn raised a prima facie case of breach of Article III (2) which required rebuttal by evidence demonstrating the contrary. In doing so, he emphasised, citing The George S [1989] 1 Lloyds Rep. 369 at 370 per Lord Donaldson MR, that proof met by counterproof may in its turn call for and be met by reinforcement of the original proof, with perhaps further iterations. It is not of course in every case that the evidential pendulum swings in this way to any real extent or even at all.

The most substantial source of evidence as to the lining (Footnote: ¹⁰) of the containers consisted of photographs taken by (a) Captain Martin Tanke, the surveyor appointed by cargo underwriters, during his inspection of (all but one of) the consignments in Germany, and (b) the stevedores at Buenaventura immediately prior to or during stuffing.

Captain Tanke’s photographs, which were attached to his survey reports, relate to twelve containers (from six different consignments). They were taken at a time when unstuffing had started, proceeded to varying degrees, and in one case been completed. The paper had been necessarily disturbed and sometimes removed in whole or in part in doing so and/or to facilitate the inspection of the cargo. In addition, adhesive tape used to attach the paper to the walls may sometimes have become detached through condensate wetting and with it part of the lining. These matters affect the extent to which these photographs can be “read back” to determine the nature and state of the lining before stuffing.

The load-port photographs, which of their nature obviously do not raise the same problem, relate to five containers (from four different consignments), three of which also feature in Captain Tanke’s photographs (Footnote: ¹¹) . All had on outturn significant to high percentages of damaged bags.

Reference was also made before me to Captain Tanke’s survey reports. Though in addition he gave oral evidence, it was clear to me that he had, quite understandably, little recollection beyond what was contained in the reports.

Perusal of the photographs, in particular those at Buenaventura, indicates that the paper was affixed to the walls with adhesive tape to a height just short of the roof (Footnote: ¹²) . In the case of the roof “lining”, the paper appears to have rested on string stretched between the two walls and was either taped to the outside of the roof at the door-end or folded down there. The result was to suspend the paper not far below the ceiling and around probably 25–30 cm above the bagged cargo (Footnote: ¹³) .

There was some discussion before me as to the gauge or weight of the Kraft paper. In two of his reports Captain Tanke referred to the containers having been lined with “thin Kraft paper”, and in his written witness statement referred to thin Kraft paper having been applied in all the containers which he inspected. He believed that the paper would have had a specification of about 80 gr/m ² . I treat this with some reserve, since in cross-examination he estimated at 125 gr/m ² a sample of paper with (in fact) a 250 gr/m ² specification – though it would also be dangerous to apply the same multiple mechanically to lower grades of paper. Two of the experts agreed in their joint statement that typical Kraft paper gauges range in weight from about 60 to 150 gr/m ² subject to local availability, though samples produced by the claimants indicate that it can be obtained in weights extending up to 440 gr/m ² . None of the expert witnesses was asked to estimate the weight of the paper as shown in the photographs. While without such assistance I could not properly arrive at any figure myself, I suspect that Captain Tanke’s estimate of 80 gr/m ² probably understates the weight, even significantly.

A specific positive complaint of the claimants at the trial was that the lining of the containers was not, but should have been, done in a double layer of paper. In response, the carrier referred to packing lists for the first two consignments which showed a cross in a box marked “doble” under “Kraft” . Spreadsheets produced by the stevedores also recorded “doble papel kraft” , though unhelpfully without any indication of the source of this information. My attention was further drawn to emails intimating a decision by the carrier by or around this time that as of 1 March 2012 all quotations for shipments of coffee beans LCL/FCL should be based on two layers of Kraft paper, with the customer to bear and be invoiced for the increased expense (Footnote: ¹⁴) . Conspicuously, however, the carrier did not produce any material showing that the shippers had in fact been invoiced for a second layer. Nor, though most of the shipments took place after 1 March 2012, was there any documentation establishing that the relevant quotations were after that date.

Moreover the photographs indicate to my eye that the paper was not “doubled” in any normal sense of that word. There was a substantial horizontal overlap between two lengths of single paper applied separately to the walls longitudinally. No suggestion that this overlap may have been regarded by the carrier or the stevedores as meriting the description of double paper was however advanced by Mr Missas, who provided a witness statement on behalf of the carrier.

The experts were agreed that doubling layers of paper could be considered as the same in practical effect as doubling its weight (Footnote: ¹⁵) , and it therefore does not need to be considered separately. Moreover, what matters is not – whether in single or double layers - the weight as such, but its effect as regards possible reduction of heat and moisture transmission. It was agreed that no weight of Kraft paper applied to the walls would materially affect conduction of heat away from the outside of the stow, since its insulating properties were negligible. That leaves weight as a factor in (a) the transmission of heat between the top of the stow and the roof and (b) the absorbency and water retention of the paper. These questions will feature later in the course of this judgment.

Condensate damage has been observed generally for many years where hygroscopic commodities – not only coffee beans but also cocoa beans, grains, and rice – are carried in steel containers (Footnote: ¹⁶) subjected to a significant fall in external ambient temperature. The physical laws and phenomena involved are far from arcane. In short, warm moist air rising by convection through and out of the top of the cargo and thence through the air-space above condenses on the colder roof of the container. Thus dried and reduced in temperature the air gains in weight, falls, and flows preferentially down the outside of the stow which has been cooled by contact with the walls. The condensate on the ceiling drips on to the top of the cargo and also runs down the walls making contact with the outside of the cargo.

The accepted temperature in Colombia at the time of loading would have been around 30 ^o C, declining little if at all before transhipment at Balboa. At that temperature the absolute humidity, i.e. the maximum water vapour which air can retain, is around 30 gr/m ³ , falling significantly with any decrease in temperature. Relative humidity is the percentage of that potential maximum in fact present in the air, and therefore increases as the temperature declines, eventually reaching 100%, i.e. saturation level, at a temperature known as the dew point. On loading, the top of the cargo and the air above interact creating a moisture equilibrium known as the equilibrium relative humidity (ERH). Given the small volume of the headspace relative to the total capacity of the container, the ERH would be almost entirely dependent on the moisture content and relative humidity of the cargo. At 30 ⁰ C a moisture content of 11% to 12% equates to around 70% relative humidity, increasing to 100% at 23.3 ^o C. Subject to an inevitable degree of approximation, that was therefore in the present case the dew-point of the air in the headspace, and condensation would accordingly occur on any surface at or below that temperature.

The figures attached by Mr John Williamson to his expert’s report show that the ambient temperature in North West Europe did not reach 23.3 ⁰ C even in May and by day and was generally well below that figure, sometimes – presumably at night - falling below zero. The steel of the container roof would have adjusted within a few hours to the ambient outside temperature. Moist air reaching the roof would therefore inevitably condense on that surface.

At the same time, the high heat capacity of coffee beans coupled with their low heat conductivity and the dimensions of the stow would have resulted in only a small heat loss through the walls from the sides or ends of bags in direct contact with them, so that the centre of the cargo would have remained close to the load temperature of 30 ^o C with its relative humidity of 70%. Condensation on the roof would therefore continue unless the convection of warm moist air through and out of the top of the stow were in some way interrupted.

The opinion of Dr Martin Jonas, the expert called by the carrier, was in essence that this process would have lasted for as long as the containers were exposed to cold ambient temperatures after arrival in North West Europe, with the stow retaining sufficient heat and moisture to fuel the process continuously for at least several weeks. This thesis was deployed in support of an argument on behalf of the carrier that no amount of Kraft paper, and however applied, would have prevented the wetting and the claim must therefore fail at least for want of causation if not at the prior stage of liability.

The contradictory theory of Dr Yves Wild, called by the claimants, was that the top layer of the cargo, cooled by convection within the airspace and radiation through it, would have declined sharply in temperature within a short time of a fall in the external temperature, and thereby halted the arrival in the airspace of moist air from the interior of the cargo. The water vapour available for condensation would from that point have been restricted to the moisture in the small airspace, by calculation a mere 0.12 litres. Condensation would have ceased thereafter until a recovery in the ambient temperature brought a repeat of the process, perhaps several times, triggered for example by an alternation between day and night temperatures. The small amount of condensation occurring temporarily in the meantime on each such occasion could, Dr Wild appeared to suggest, be absorbed by the Kraft paper and would therefore never impact on the top or of the cargo. On this basis, the claimants contended that the damage could only have resulted from some inadequacy in the quality or thickness of the lining paper or in its deployment.

Dr Wild expounds his theory at some length in paragraph 3.3 to 3.4 of his first report. At its heart lies a computer generated model created by him of temperature changes in the cargo in response to falls in ambient temperature. For this purpose the cargo was notionally divided into 2,000 cuboid elements approximately 24 x 31 x 28 cm. The change in temperature in each cube was recalculated at the elapse of every 5 minutes for periods of 12, 24, and 96 hours, assuming an initial internal temperature of 30 ⁰ C throughout the cargo and an external air temperature plunging instantly at the outset to 5 ⁰ C. The model showed that there was almost no change at the centre of the stow even after 96 hours, with a gradient down to around 17 ⁰ C in the top 24 cm after as little as 12 hours, 11 ⁰ C after 24 hours, and 6 ⁰ C after 96 hours (and even lower temperatures in the top corners).

The heat diffusion through the cargo in Dr Wild’s model was calculated solely on the basis of conduction. This completely excluded the effect on the top layer of upward transportation of heat by convection through the cargo, around 45% of which consists of interstitial air.

The model also proceeded on the assumption that the cargo filled the entirety of the container and lost heat by conduction through all six surfaces (walls, floor, and roof) with which the cargo was on this hypothesis in complete contact. This, as Dr Wild recognized, departed significantly from reality by ignoring the airspace above the cargo where instead of conduction the heat transportation would operate by radiation and/or convection and be potentially affected by the interposition of Kraft paper. However, “in order not to increase the complexity of the calculation too much” he decided to use for the top of the cargo a reduced heat transfer coefficient of 3.5 W/m ² K, instead of the 6.5 W/m ² K he used for the sides. In his oral evidence, he proved unable to identify any real basis for adopting this figure. In his written report he went into the question to some extent, observing that the appropriate coefficient would depend almost entirely on the heat transfer between air and paper (and vice versa), for which he used a transfer coefficient of 5.0 W/m ² K on the assumption of paper of 80gr/m ² with a thickness of 0.1mm. His discussion led him to the conclusion that the exact overall impact of the use of Kraft paper “ cannot be easily calculated due to lacking knowledge of exact framework conditions and empiric data” , but he felt able to add that “[i]n general, it can be stated that even the use of very thin Kraft paper leads to significant thermal insulation between cargo and top air space inside the container”. In these circumstances, counsel for the carrier had good reason to characterise Dr Wild’s figure of 3.5 W/m ² K as arbitrary, though I would prefer to describe it as lacking any tangible foundation. It is also based on 80 gr/m ² and 0.1 mm, which may well be significantly less than the true weight and thickness. Dr Wild did not run his model to reflect any greater reduction in the coefficient, and I have no basis to conclude that any such re-run would not produce materially different results.

Taken separately or together, these factors precluded Dr Wild’s model from establishing his desired conclusion that the heat removed from the top of the cargo by convection and radiation would have exceeded the heat arriving from below, let alone the extent and rate of any suggested excess.

His concern with the temperature at the top of the stow related to a subsidiary consequence. The hygroscopic capacity of coffee beans increases with a decline in temperature. It follows that some of the moisture in the air rising by convection through the beans would be absorbed by them and removed before the air left the stow, correspondingly cooled, at the top. What matters however is the rate and scale of this phenomenon. Dr Jonas told me that it would be slow. Dr Wild’s opinion was that the amount of water absorbed would be so small as to scarcely affect the moisture content of the beans, but neither he nor Dr Jonas quantified the amount involved and correspondingly removed from the convected air. I have therefore no basis on which I might properly find that the reduction in its dew point would have been such as to preclude further condensation on the ceiling of the container.

Significantly, also, if the top of the cargo had fallen below the dew point of around 23.3 ⁰ C relevant to the air coming from its centre, the moisture in that air would have wet the beans directly. That would have produced a crust of mould quite unlike the superficial damage caused by drips. No such crust was observed on the surveys. Dr Wild’s theory in this respect collided with empirical fact.

Temperatures at or below 5 ⁰ C were recorded (most likely at night) at Rotterdam, Hamburg and Bremen for most of the consignments, and though coupled with highs ranging upwards from 10 ⁰ C reached 21 ⁰ C only for two consignments and are never reported to have exceeded 23.3 ⁰ C. Accordingly, the temperature of the ceiling would have been below that figure, for most of the time very significantly so, which would have provoked condensation even if the absolute humidity had been noticeably reduced by Dr Wild’s suggested process.

In the light of all these matters, I am unable to accept Dr Wild’s theory and find that his model does not assist in explaining what happened in the containers in the present case. To the contrary, I accept Dr Jonas’ analysis and explanation as essentially correct: the condensate was generated, not in a series of abbreviated episodes as Dr Wild suggests, but by a process which would have continued without interruption as long as the ambient temperature and thus the ceiling of the containers was below around 23.3 ⁰ C, as would have been the case once they had arrived in North West Europe.

The carrier’s case on the inevitability of damage rested on Dr Jonas’ opinion that a thicker grade of Kraft paper, or multiple layers of Kraft paper, would not have made any significant difference to the extent of cargo wetting found on outturn. He stated in his report that:

It is self-evident that once paper is saturated water will pass to the cargo, either directly in the case of the wall lining or by drips in the case of the paper suspended under the roof. What matters for present purposes is rather the capacity of various grades or types of paper or card to absorb and retain moisture and hence the time which would be required to reach saturation. On that key question Dr Jonas gave no assistance. Some information was provided by Dr Wild. He told me in his report that Kraft paper of 150 gr/m ² may hold up to 250 gr of water per m ² , so that the roof lining with a surface area of about 14 m ² could absorb 3.5 litres of water. On the basis of an airspace between about 4 and 7 m ³ and a consequent condensation rate of 122 gr (= 0.12 litres) of water per hour, saturation would therefore occur after 29 hours. These figures assume however that all the condensation would drip on to the suspended paper above the cargo, whereas experience in the present cases suggests that a significant quantity would run down the walls and thus encounter a further mass of paper. On the other hand they also assume that the wetting of the paper would occur evenly, ignoring any local saturation. They are also given only for paper of 150 gr/m ² : whether or not I might incline to the assumption that the absorbency of Kraft paper varies directly with weight, though this may depend on the quantity of glue which it contains and its effect on the shininess of the paper’s surface, I would certainly require expert assistance before accepting that a comparison between paper and corrugated card would show any similar linear correlation.

I am therefore not in a position to determine how long it would take to saturate paper or card of varying weights and characteristics, or whether it would occur within the contemplated duration of the contracted transport. Even assuming that despite deploying a maximum weight and/or using corrugated paper or card saturation and water transmission would arrive eventually, the carrier has not established that this would have happened within the time-scale of the contracted carriage.

The proposition that damage during such carriage cannot be prevented by any form of paper or card lining must also confront historical experience. If correct, one might have expected it to have engendered long since and still continuing an industry-wide spate of claims and litigation, of which the carrier produced no evidence (Footnote: ¹⁷) . The carrier’s own announcement of its change to double layers in March 2012 would also appear to have been pointless if its contention in this court as to inevitability of damage were correct. Its subsequent experience might also have been expected to cast light on this question, but the carrier gave no evidence as to this, or indeed its experience before 2012, and I am thus without any assistance from this source.

In short, the carrier has not made good its contention that no lining with paper or card could have prevented the damage in the present case.

Moreover, even if that contention had been correct, it would have availed the carrier nothing.

A carrier is obliged under Article III (2) to employ a sound system. Since the “soundness” relates to the prevention of damage to a normal cargo from the risks reasonably to be expected during the contracted carriage, it is no answer to an allegation of breach of that obligation to say that, since such damage is always unavoidable, no such system exists.

If the argument is advanced alternatively as one of causation, the suggested non-existence of any sound system would entail that the counterfactual required to establish (and then quantify) damage and loss would be devoid of content.

Despite its argument on the inevitability of damage, the carrier also submitted (inconsistently, as it seems to me) that the lining was in accordance with a sound system. While it would be inappropriate to attempt a comprehensive definition of these words, not least because they are only a judicial gloss on the text of Article III (2), the concept of a sound system (and the base word “properly”) must in my view require as a minimum that there exists a rational, adequate and reliable basis for concluding that it will prevent the otherwise threatened damage.

As I pointed out earlier, while a sound system is not required to take account of the peculiarities of an individual cargo, these coffee beans were entirely typical of an extensively carried commodity. The propensity of all bagged coffee beans to generate moisture and potentially condensation during carriage in containers to colder climates was well-known in the industry, and the lining of containers by the carrier was clearly in reaction to, and recognition of, the risk of damage from this source. It had therefore to demonstrate a rational, adequate and reliable basis for concluding that its lining arrangements could be expected to prevent such damage from occurring during the contracted carriage (unless disturbed or affected by one of the true excepted perils in Article IV (2) such as Act of God or strikes).

Conspicuously lacking from the evidence of the carrier was both any identification of the weight of the paper used in these containers (whether or not, as it contended, in a double layer) and any explanation of why that weight and type had been chosen. The only significant way in which paper or card might assist in preventing damage, as the expert evidence confirms, would be by moisture absorption and retention. Ideally, one would therefore seek to ascertain the amount of condensate which could be expected in the course of such carriage to North West European destinations and the time before saturation of paper or card of varying weights or types. Alternatively, particularly if this cannot in practice be properly ascertained by theoretical calculation, a suitable empirical study might establish that a particular weight and/or type of paper or card is sufficient in practice to prevent damage throughout such carriage. No such calculation or study was referred to in evidence, either instigated or effected by the carrier itself or produced or endorsed by an outside body such as a trade association or P & I club. It may well be that such work has never been done by anyone. In the absence of such material, however, the court is without any basis on which it might permissibly find that the use of any particular weight or quality of lining paper or card (including whatever may have been deployed in the present case) would have constituted a sound system.

A number of publications, some of a certain antiquity, referring to the use of Kraft paper or card, were suggested by counsel for the carrier to indicate a general practice to which it had conformed. However, they showed no consistency as to doubling, and none specified a weight or thickness, the key parameters so far as prevention is concerned. I am unable to regard these documents as recording or even evidencing any general practice in this regard, let alone one operative in 2012. In any event, a general practice – had one existed - could not itself have rendered a system sound in the absence of any appropriate theoretical or empirical underpinning, and I do not read any judicial pronouncement as having decided or even suggested the contrary (Footnote: ¹⁸) .

In short, the carrier has failed to establish that it adopted a sound system.

A decision to the contrary would moreover not have ended the difficulties in the carrier’s argument. If the system was a proper one, why did it fail to prevent damage in all but two of the twenty containers? The remaining possibility, that the system was sound but had been incorrectly implemented (Footnote: ¹⁹) , would have required the carrier to show that this was not due to lack of care. I do not see how on the material before me it could have succeeded in doing so.

In view of my conclusions so far, it is unnecessary for me to decide this residual issue; and I will deal with it more shortly than might otherwise have been the case.

The photographs record a variety of stow configurations, sometimes mixed or even – to use the expression of Captain Tanke - “chaotic”. Originally, the claimants contended that all the bags should have been loaded in saddle stow, a configuration in which alternate layers are shifted laterally (as in brick-laying), on the basis that this would reduce convection by closing off air channels. In the course of the trial it became however clear that given the degree of floppiness of the bags almost any configuration would have this effect, and that it had been equally well achieved in all the stows shown in the photographs.

The claimants’ submission then shifted to an argument that all the bags should have been stowed longitudinally, i.e. lengthwise in relation to the side-walls of the container. Though this was said to be supported by their own expert, Mr Williamson, his view, stated clearly in his report, was in favour of transverse stowage; indeed he criticised the stow in most of the photographed containers precisely because it was lengthwise. His reasoning was that crosswise stow exposed only the ends of the bags to contact with the side walls, albeit to some extent counterbalanced by the fact that the long sides were then in contact with the far wall and possibly the doors, and therefore minimised the quantity of beans affected.

In attempted justification of their contrary stance, the claimants observe that longitudinal stowing reduces the number of bags in contact with the side walls, though they fail to allow for the correlative (albeit lesser) increase in the number of bags in contact with the far wall and the doors. On behalf of the carrier it is pointed out that a bag pressed lengthwise against a side-wall has a greater surface exposed to potential wetting and therefore risks damage to a greater number of beans. Against this, it is argued on behalf of the claimants, account must be taken of the fact that some of the potential financial loss would relate to the cost of inspecting and sorting any bags showing signs of wetting (including stains), rather than directly to the condemnation or rejection of damaged beans.

Overall, I am unpersuaded that the balance of advantage lies objectively with a longitudinal stow, let alone with a degree of uncontroversial clarity which should have led any prudent carrier to opt for it. I am also more inclined to accept the opinion of the claimants’ expert than the opposed argument which their counsel now urges upon me. Indeed, I am far from clear that they are forensically free to disown his evidence in this way.

The claimants sought permission to amend the Particulars of Claim by adding an allegation that the periods spent at Rotterdam and Hamburg, during which the condensation would in at least large part have occurred, constituted a breach of the obligation to prosecute the voyage with reasonable despatch. This was a discrete new allegation which should have been made much earlier and supported by clear and targeted evidence (almost certainly from experts) as to generally accepted periods for transhipment in carriage of this sort. I indicated as much in the course of argument by counsel, and now formally reject the application. In any event, given my previous conclusions, the claimants’ proposed allegation, even if proved, would have made no difference to the outcome of the action.

There will therefore be judgment for the now agreed sum of US $ 62,500 plus interest. The parties are asked to determine how this should be apportioned between the claimants and to draft an order accordingly.