American Overseas Marine Corp v Golar Commodities Ltd

The claimants, American Overseas Marine Corporation (“AOM”), are the managing owners of a liquefied natural gas (“LNG”) carrier, the “LNG Gemini”. By a time charterparty of 3 January 2011 they chartered her to the defendants, Golar Commodities Limited (“Golar”). Clause 30 of the charterparty was headed “Injurious Cargoes” and it provided:

“No acids, explosives or cargoes injurious to the Vessel shall be shipped and without prejudice to the foregoing any damage to the Vessel caused by the shipment of any such cargo, and the time taken to repair such damage, shall be for Charterers account”.

AOM allege that in May 2011 Golar loaded at the Cameron Terminal, Louisiana a cargo of LNG that was injurious to the vessel in that it contained debris, in particular metal particles, and that as a result “major repairs” to the ship were required after she was dry-docked on 24 October 2011 at the Keppel Subic Shipyard (“KSS”) at Subic Bay in the Philippines and her cargo pumps and tanks were found to be contaminated. Relying on clause 30, they claim damages of $1,933,933 for the cost of repairs and time lost and an indemnity for this sum and against any claims for contamination of other cargoes.

Golar deny the claim, and bring a counterclaim for $618,060.23 (plus statutory interest) in respect of the vessel and a charter of a sister ship, the “LNG Leo”. Neither the counterclaim nor Golar’s right to set it off against the claim is disputed.

Golar accepted before me (although not in their pleaded defence) that, when the vessel loaded at the Cameron Terminal, metallic particles and other debris was trapped in the manifold strainers and loading arms, and that some debris (to which I refer as “Cameron debris”) was probably carried into the cargo tanks. The main issues are:

The trial took three days. AOM’s only witness of fact was Mr Brian Cain, who was their Director of Engineering at the relevant time before he retired in September 2012. They also relied on statements of three witnesses whom Golar did not require attend for cross-examination (reserving the right to rely on discrepancies between the statements and documents in evidence): Captain Donatan Balog, the Master of the “LNG Gemini” when she loaded at Cameron, Mr Marin Magazinovic, the Chief Officer, and Mr Mario Marković, the Cargo Engineer. Golar had three witnesses of fact: Mr Tom Christiansen, who at the relevant time was their Vice President, Operations, Mr Arvid Nygård, their Fleet Manager, and Mr Bo Johansson, who is the Managing Director of Viking Engineering Pte Ltd (“Viking”).

I consider the evidence of the defendants’ witnesses reliable. I was less impressed by Mr Cain’s evidence, which was shown in cross-examination to be unreliable in some respects. By way of example:

Each party called expert evidence of an engineer. AOM called Mr Kenneth Shortall, a consultant with TMC (Marine Consultants) Ltd and a marine engineer. He has worked for Shell and served as Chief Engineer on oil and gas tankers. Golar called Mr Christopher Clucas of CWA International, a chartered engineer, who is Group Fleet Director of Bernhard Schulte Shipmanagement Isle of Man (known as “Dorchester Marine”) and has particular expertise with regard to both LNG and Liquefied Petroleum Gas (“LPG”) carriers. Both were well qualified to give evidence, but Mr Clucas’ experience gave his evidence more authority, and, no doubt because of his practical experience, his views were more realistic.

In his second report Mr Shortall referred to what he had been told by a “colleague”, Mr Tony Grainger, about the KSS. AOM served notice under the Civil Evidence Act, 1995 that they did not intend to call Mr Grainger as a witness because the costs of doing so would be disproportionate. Golar objected to this evidence on the basis that it was expert evidence for which permission had not been sought or granted. In the event, AOM did not apply for permission to adduce the evidence.

Mr Nigel Eaton QC, who represented Golar, rightly observed in his opening submissions that the statements of witnesses of fact expressed opinions that were admissible only as expert evidence, and said that he would not cross-examine Mr Cain about them. Equally he might have said that the expert witnesses gave their views on the facts. I do not say this in order to criticise the witnesses: in this sort of case, the line between factual matters and questions of expertise is often blurred. But I welcome Mr Eaton’s approach, and I discouraged counsel from questioning the witnesses of fact on their opinions based on their general experience.

The parties put before the court agreed bundles, which were admissible as evidence of their contents (under CPR 32PD para 27.2). AOM also sought to rely on two other files of documents. These comprised documents that they had disclosed late (I understand only a fortnight or so before the trial). One file (file M) contained detailed results (extending to nearly 300 pages) of an analysis in 2014 by SGS of sediment said to have been taken from the vessel in October 2011. Golar had served a notice under CPR 32.19 that they required the authenticity of the documents to be proved, and they were not. Mr Nicholas Saunders, representing AOM, submitted that it would be “just” to admit the results of SGS’s analysis, but did not identify any basis on which they might be adduced. I do not accept them as evidence.

The other file (file L) contained eighteen documents, most of which were dated 2011. Mr Saunders sought to rely in particular on (what purport to be) Heads of Agreement for the Charter of the “LNG Leo” made in November 2011, a charter of the “LNG Gemini” dated 12 August 2011 and a charter of her dated 12 December 2011. AOM plead a claim for loss of use of the “LNG Gemini” based on a daily hire rate of $55,000. Mr Cain explained that this hire rate was used on the instructions of Mr Thom Merrell, the President of AOM, and sought in his third statement to justify it by reference to the charters.

On 9 May 2013 Teare J had ordered that statements of witnesses of fact and notices in respect of hearsay evidence be exchanged by 13.00 on 6 September 2013. AOM served no notice under the Civil Evidence Act 1995 about what Mr Merrell had said. Their witness statements did not deal with the quantum of their loss until late on 14 February 2014 (the Friday before the trial was due to start on the following Monday) AOM delivered to Golar a third witness statement of Mr Cain. It exhibited (several versions of) two invoices that had first been disclosed on 13 February 2014, and it explained an internal memorandum dated 11 January 2012 that Mr Cain had sent to Mr Merrell. In the end Mr Eaton did not resist much of the statement being adduced because it simply provided a guide through documents that were already in evidence. However, he objected to a section headed “Loss of use”, complaining that, because the documents in file L were disclosed late and Mr Cain’s third statement was served late, Golar had not had a proper opportunity to assess the evidence, to take advice from a broker about the market rate and, if so advised, to apply for permission to call evidence about it. I upheld those objections. AOM gave no reason for the late disclosure, and there was no acceptable reason that evidence about quantum was not included in an earlier witness statement: Mr Saunders said that “It was just something that hadn’t been done until that time”. Golar were prejudiced by AOM’s failures. I therefore did not receive the documents in file L as evidence and did not permit AOM to introduce the controversial section of Mr Cain’s third statement. This put AOM in difficulties about proving the claim for loss of hire (which represents about half of their financial claim), and Mr Saunders applied for a separate trial of quantum. I refused the application: there was no reason to defer all questions of quantum. However, I said that, if AOM sought an order for the assessment of damages for loss of time, I would decide in this judgment whether to order this.

The “LNG Gemini” is a 100,000 mt “Aquarius” LNG carrier, capable of carrying about 125,000 cu m of cargo. She was built by General Dynamics, AOM’s parent company, in 1978 at the Quincy Shipyard in Massachusetts. She is one of ten sisterships, and AOM are the managing owners of three of them, including the “LNG Gemini”. Her registered owner is Patriot I Shipping Co (Patriot”), Pronav Management GmbH & Co KG (“Pronav”) are her technical managers, and she is classed by the American Bureau of Shipping (“ABS”).

The vessel has five cargo holds, and in each is a spherical, aluminium, insulated tank, called a “Moss” tank. Before LNG is loaded, air must be removed from the tanks because LNG is combustible if mixed with oxygen. It is forced out by filling the tanks with inert gas, which has passed through a drier to remove water vapour: the cold LNG might freeze any residual water and prevent valves and other equipment from operating. The inert gas is then forced out by warm natural gas, and the tanks are cooled before cargo is loaded so as to protect the tank surfaces from a thermal shock from the cold cargo. (LNG is carried at temperatures of about -150° C.)

LNG cargoes are loaded from shore tanks through loading arms and the ship’s manifolds. Conventionally (since the much-publicised problems encountered by the “Hilli” on her maiden voyage in the 1970s) manifolds are fitted with conical wire-mesh filters to stop debris in the LNG from entering the tanks. They are usually 20 mesh or 60 mesh filters (filters with 20 or 60 meshes per inch). Guidelines published by the Society of International Gas Tanker and Terminal Operators (“SIGTTO”), first in 1984, make this recommendation:

“Protection from contamination may be required particularly following the entry into service of an export terminal or of a ship or following any maintenance or refitting in an export terminal or on a ship which could affect the internal cleanliness of the terminal‘s storage/loading systems or of the ship’s cargo system respectively. Whenever such conditions apply, strainers of ASTM 60 mesh (nominal aperture of 0.25mm) should be installed on shore, on ships or on both in accordance with the accepted philosophy, during loading and discharge of the first two subsequent cargoes from a terminal to which the conditions apply or the first two subsequent cargoes carried by a ship to which the conditions apply.

Thereafter, provided that all strainers at both loading and discharge are exhibiting clean condition on inspection following cargo transfer and provided that there is to be no significant increase in cargo transfer flow rate, the ship’s master, in consultation and agreement with the receiving terminal, may replace the shipboard strainers with strainers of ASTM 20 mesh (nominal aperture of 0.84 mm).”

In July 2011 (after the “LNG Gemini” had loaded at the Cameron Terminal) SIGTTO published “Manifold Recommendations for Liquefied Gas Carriers”, which confirmed that:

“For LNG, current practice is to use either ASTM 20 mesh or ASTM 60 mesh (nominal aperture 0.25mm and 0.84mm), the choice tending to be dictated by commercial rather than strict technical reasons. If mesh finer than ASTM 60 is desired, head loss should be carefully considered.”

The manifolds in the “LNG Gemini” had unusually fine 100 mesh filters: Mr Shortall had never come across such fine filters on an LNG carrier before. According to Mr Clucas’s unchallenged evidence, their nominal aperture was 0.149 mm, that is to say they would block particles of more than 0.149 mm.

In each tank the vessel had two cargo pumps (used to load and unload cargo) of a design commonly used in LNG carriers, and also in LPG carriers and on-shore installations. They are called “Carter” pumps: they were originally designed and manufactured by J C Carter, who made the pumps in the Aquarius vessels, but pumps of this kind are now generically referred to as Carter pumps regardless of the manufacturer. J C Carter have been taken over by Atlas Copco: the exact legal relationship between J C Carter and Atlas Copco is not clear from the evidence, and I simply refer to Atlas Copco because the distinction between them is not important. The design of Carter pumps is unusual in that the electrical motor is coupled directly to the impellor, and so they are near to the floor of the tank at the bottom of a “pipetower”, which runs the depth of the tank and carries piping. They are therefore submerged in liquefied gas when the tank is loaded: liquefied gas passes through the motor when the pumps are running, and cools and lubricates the bearings. The pumps on the “LNG Gemini” had formerly been on a sistership, the “LNG Capricorn” and they were transferred in 2007. Atlas Copco had overhauled them in 2008.

Each tank also has a spray pump (or cool-down pump). These are also Carter pumps, but they are much smaller than the cargo pumps. Cargo pumps are usually about 1 or 1.5 metres tall, whereas the height of spray pumps is about 50cm. The spray pumps are, as Mr Clucas explained, “quite near the bottom of the tank”. When a cargo is discharged, a “heel” of LNG is usually kept in the tanks and used to spray them during a ballast voyage to keep them cool. Spray pumps are usually protected by filters, which are often quite coarse: the nozzles of the sprays are typically some 18 to 25 mm. The spray pumps on the “LNG Gemini” have (or at the relevant time had) 150 mesh filters (about 0.1 mm), which again are unusually fine.

The tanks have automatic safety relief valves to release excessive pressure by venting gas into the atmosphere, and also manually operated relief valves.

The “LNG Gemini” records and monitors the cargo in the tanks through sensitive transducers in the tanks, which send the Cargo Control Room readings, including readings of the level of cargo in the tanks, its temperature and pressure inside the tanks.

From when she was delivered as a newbuild, the “LNG Gemini” was on a long bareboat charter, latterly to Burmah Gas Transport of Stamford, Connecticut, until 29 January 2011. She and her sisterships, “LNG Leo” and “LNG Virgo”, were re-delivered in ballast at Subic Bay, and Mr Cain supervised her re-delivery survey. The survey was focused on her structural condition, and, although the cargo holds were empty, because of time constraints Mr Cain and other inspectors went into only hold no 3 and the pumps were not opened for inspection. Mr Cain signed a report that the vessel’s condition was “satisfactory”, that is to say (according to the definitions in the report form) “at condition of wear and tear of such an extent as not to essentially affect the original strength and/or efficiency”.

From 25 January 2011 two surveyors from Lloyd’s Register carried out a Condition Assessment Programme (“CAP”) survey of the vessel’s hull. Their report, dated 18 May 2011, said under the heading “Cargo Tanks/Cargo Holds” that “the cargo tank structures were all recorded in good condition with no substantial corrosion”. I do not understand from this that there was minor corrosion, and Mr Eaton did not suggest that I should do so. Mr Shortall referred to a red colour that he discerned on a photograph taken of the bottom of one of the tanks during the survey, but the photograph is unclear and to my mind is not evidence of rust.

Mr Cain’s evidence was that the surveyors carried out a “close-up” inspection of the cargo tanks and ballast tanks, that he was with them when they inspected two of the tanks and that he saw no debris in any tank. However, the Lloyd’s Register report shows that, while the ballast tanks were inspected close-up, the cargo tanks were simply covered in a general examination of other spaces of the vessel. This is readily understandable: it was a hull structural survey, and the surveyors’ concern was, no doubt, directed to any signs of cracks, corrosion, thinning steel or other structural deterioration. I do not accept that the cargo tanks were inspected closely.

Mr Shortall referred in his report to photographs said to have been taken by “the superintendent” on 27 January 2011 and to show the cargo tanks without contamination. He said in cross-examination that he was provided by AOM’s solicitors, Messrs Hill Dickinson, with four such photographs, but there is no evidence about who was “the superintendent” who took them, why they were taken or how those sent to Mr Shortall were selected. I cannot draw any inference from them about the general condition of the tanks. In all there are in evidence eight photographs dated 27 January 2011, but the additional ones add nothing to those sent to Mr Shortall.

On 1 and 2 March 2011 the vessel underwent an ABS class survey at Subic Bay. It recommended that specified matters, mostly or wholly by way of steelwork and the ballast tanks, be dealt with by 28 or 30 November 2011. There is no evidence that the tanks were cleaned at Subic Bay. Mr Shortall and Mr Clucas agreed in the memorandum of their meeting that her cargo tanks were clean when she departed from Subic Bay, but, as Mr Clucas made clear, they were simply acknowledging what was in the CAP report, and they had no more information about the state of the vessel.

In January 2011 Golar had entered into the charterparty of the vessel for 60 days (15 days more or less and with two charterers’ options to extend the hire for further such periods). The rate of hire was $28,000 per day. She was delivered into the charter on 1 March 2011. Her tanks were inert. On 2 March 2011 Captain Balog joined her as Master. He is a Croatian and had served on LNG carriers for some years: in 2003 he trained as a Chief Officer on the “LNG Aquarius”, a sistership of the “LNG Gemini”, and served on her in 2004; he served as Chief Officer on the “LNG Gemini” between 2004 and 2006; and he was promoted to Master in 2006, and has captained various of the sisterships without any previous problems of any relevance.

On 7 March 2011 Golar ordered the vessel to carry a full cargo of LNG from the Cameron Terminal to Senboko, Japan. The terminal, operated by Sempra LNG (“Sempra”), was primarily used for importing LNG, but had started to export it. The “LNG Gemini” arrived there on 30 April 2011, and at 22.29 was all fast at the Cameron North Terminal. By 07.30 on 2 May 2011 she had completed “gassing up”, whereby the inert gas was forced out of her tanks, and she was ready to cool her tanks in order to receive cargo. During this process, LNG evaporates, and the evaporated gas (gaseous natural gas, or “GNG”) is usually returned to the exporting terminal. However, the Cameron Terminal advised that it could not take back all the GNG that would be produced, and the vessel’s boilers could not burn it off. It was therefore decided that she should load only LNG to cool tank no 3, and on 5 May 2011 she loaded it slowly so as to control the GNG. The crew inspected the strainers after this parcel was loaded and found some debris but “no more than was considered usual”, as Mr Shortall put it and I accept. Then at 16.50 on 5 May 2011 the “LNG Gemini” sailed from the North Terminal in order to cool the other four tanks at sea with LNG from hold no 3. These procedures were completed by 07.50 on 8 May 2011. Te

The “LNG Gemini” then went to the Cameron South Terminal to load her cargo, and she was all fast by 13.40 on 9 May 2011. Pre-arrival checks were completed, and Mr Magazinovic and Mr Marković reported that the manifold filters were free of debris. She loaded from tank 201, one of the Terminal’s three shore tanks. Loading began at 19.29 on 9 May 2011 through manifolds nos 39 and 43. The flow was to be controlled at less than 1,000 cu m per hour, below the more usual 10,000 cu m per hour. The valves were opened slowly, initially at 20% and gradually increasing to 100% by 20.28.

At 20.40, the Third Officer, who had been stationed at the manifolds with instructions to check the differential pressure, reported that it had increased to about 50 pounds per square inch (60/10 PSIG). This was well within the design limit of the strainers, which were for differential pressures of up to 150 psi, but it was unusually high: Mr Shortall said that it would normally be about 2 to 4 psi, and according to SIGTTO’s 2011 publication the maximum pressure drop across dual directional flow strainers in clean condition should not exceed 0.5 bar (or 7.25psi). Captain Balog was cautious because the Terminal had little experience of loading LNG, and, after discussion with a Terminal representative, he decided to stop loading at 20.50. By then about 380 cu m of LNG had been loaded.

The loading arms, first arm no 39 and then arm no 43, were disconnected and pulled from the spool pieces. Sediment and debris was found at the bottom of the spool pieces and on the shore side in the filters. The strainers were removed, samples were taken and the strainers were cleaned. The strainers of two lines that had not been used for loading, nos 41 and 45, were inspected, and they were not contaminated. Captain Balog also inspected the strainer in the loading line: it was coarse (“more of a sturdy metal grating than a strainer”, as Captain Balog described it), and it held no debris.

After arm no 39had been disconnected, the by-pass valve was opened in order for the arm to be drained more quickly. This was done at 21.35, some 30 to 45 minutes after pumping had stopped. The vessel started loading again through manifold no 39 at 14.55 on 10 May 2011, but at 15.40 the differential pressure was some 20 pounds psi (35/15 PSIG), and Captain Balog again ordered that it be stopped. In total some 613 cu m of LNG had been loaded (that is to say, a further 233 cu m had been loaded after pumping had resumed). The loading arm was disconnected, and more debris was found on the shore side of the strainer. Consideration was given to moving the vessel to the North Terminal, but this was thought to be too disruptive. Loading from the South Terminal resumed at 00.19 on 11 May 2011, but a different landline was used, and apparently the cargo was pumped from a different tank, tank no 203. (The evidence is contradictory about whether loading was also resumed from the original shore-tank, no 201, but nothing turns on that.) The cargo was fully loaded without further stoppage at 13.13 on 13 May 2011.

At about 06.35 on 13 May 2011 both the Terminal’s vapour return blowers had tripped, and, because the GNG could not be returned to the Terminal, pressure inside the tanks rose rapidly. Captain Balog released it by opening a valve so as to avoid an uncontrolled vent through the safety relief valves. There is no evidence that the safety relief valves opened, and AOM do not suggest that they did. As Mr Clucas explained, and I accept, this would have made considerable noise and vapour would have issued from the vent masts and there is no evidence of this.

Mr Cain’s evidence was that he learned that “the automatic relief valves must have come close to being forced open”. To my mind, the evidence does not support this inference, and Mr Shortall could not cogently support his professed opinion that “it is highly likely that the pilot valves may have started to open”. In his letter of protest, Captain Balog said “sudden high pressure had been observed rising from 16.80 psg to 17.77 psig abs in two minutes”, but, as AOM accepted, that cannot be right. Mr Shortall pointed out that the relief valves to operate at cargo tank pressure gauges of 3.2 +/- 0.1 psig, and that the deck log book for 13 May 2011 records atmospheric pressure of 1011 mBar, the equivalent of 14.66 psi. On this basis, he suggested that the pressure in the tank was 3.11 psi (the difference between 17.77 and 14.66), and the error in the letter of protest was to refer to 17.77 psig rather than 17.77 psia. This is ingenious, but it is speculation: Captain Balog gave a witness statement for AOM, he could easily have given this explanation for his letter of protest, and he did not do so.

London Offshore Consultants Inc (“LOC”) attended at Cameron Terminal on 12 and 13 May 2011 to conduct a survey and prepare a report for the vessel’s P&I insurers. They recognised that sediment might have passed though the filters and entered the tanks, and observed (i) that it might clog the cooling sprayer nozzles during cooling operations on the voyage, and (ii) that, while there should be no problems with discharge, the terminal would probably detect sediment particles if they analysed the discharged LNG. Braemar conducted a survey on 13 May 2011 and produced a report on behalf of the vessel’s underwriters, and commented that “as far as could be ascertained, no obvious signs of external damage were noted”.

At 00.30 on 14 May 2011, the “LNG Gemini” sailed for Senboko. Later that day, after the manifolds had been defrosted, the strainers were inspected and more debris was found: according to Captain Balog there was a “considerable amount” of it, and it “appeared to be the same colour and composition as the debris removed from the strainers during the loading process”: he described the colour as “reddish-brown”. On 27 May 2011, the spray pump was removed from inside the cargo piping system, and debris was collected.

The vessel arrived at Senboko on 27 June 2011. On the voyage there was no sign of damage to the pumps or to the cargo tank relief valves or otherwise to the CCMS. She discharged her cargo without incident through 60 mesh strainers, and it was reported that the strainers “were found clean”. On 27 June 2011 a SIRE (Ship Investigation Report) inspection was conducted while the vessel was at Senboko. She then sailed in ballast for Singapore where she arrived on 1 July 2011, and on the voyage an ABS inspector inspected the cargo system and machinery. In a report dated 6 July 2011 the ABS recommended the due dates for inspecting of certain items, including overall survey of two cargo tanks, and for operational tests for and examination of relief valves, be extended to 24 August 2011. There was heel in the tanks when the vessel was inspected at Senboko and on the voyage to Singapore, and the inspectors could not go into them, but apparently they had no concerns about debris or that the pumps or any other equipment might have been contaminated or damaged when cargo was loaded at Cameron.

By 30 June 2011 Captain Baricevic had taken over from Captain Balog, and he reported to Pronav that the spray pump in tank no 5 had become clogged with what was described as “fine, moist, dusty debris”. On 1 July 2011 he reported that the strainers in no 1 tank had become similarly clogged, but that there was less debris and it was not moist. At first tests on debris from both strainers found no iron, but, after more debris was collected from the strainers in July 2011, further magnetic tests detected it.

The “LNG Gemini” proceeded from Singapore to West Africa in ballast, using spray pumps to keep the cargo tanks cool. She still had a heel of cargo, distributed more or less evenly through the five tanks, so that the tanks were cool for her next cargo. On several occasions debris was found in spray pump strainers.

On 29 July 2011 Mr Vili Vrenko of Pronav had sent to Mr Cain a report about “sediments in the LNG [Gemini] cargo tanks, which is still present”, commenting that the strainers were most recently clogged on 28 July 2011 and had been cleaned. He wrote:

“- There is a certain quantity of the sediments which are actually present in the cargo tanks. These sediments penetrated 100-mesh strainer during loading operation at Cameron and are supposed to be very fine structure.

- These fine grains are acting like abrasive inside Main Cargo Pumps/Cool Down Pump bearings, as well as inside the windings of the electrical motors.

- At this point it is unknown if these sediments can influence readings and operation to CTM – however knowing that fine metallic dust is present in the samples, the operation of capacitance sensors may be easily affected.

- [Pronav] strongly recommends that ALL of pumps inside cargo tanks are completely disassembled at the first scheduled dry-dock, bearing replaced, windings cleaned and (if necessary) re-varnished and baked.”

Pronav recommended this after discussions with Atlas Copco. AOM, however, decided against immediate work on the pumps, but to wait until the vessel next went into dry-dock.

The ABS surveyed the vessel in early August 2011 at Walvis Bay in Namibia, and they extended the date of her underwater survey in lieu of dry-docking (or “UWILD”) to 1 November 2011. She departed Walvis Bay on 4 August 2011, and proceeded to the LNG terminal at Bonny, Nigeria, where she arrived on 25 August 2011. On 10 August 2011 the pressure at the nozzles in tank no 5 dropped, and the filter in no 5 tank was found to be clogged with “the similar, very fine, powder-like sediment as in previous occasions”, but this time a magnet testing indicated that it did not contain iron. Apparently no analysis other than a magnet test was carried out on this or any other occasions that filters clogged.

The vessel loaded a cargo of LNG at Bonny for carriage to Tobata, Japan, loading being completed without incident on 28 August 2011. According to Mr Magazinovic, a “small amount of debris was found in the manifold strainers, but no more than [they] would usually expect …”. It was reported by Pronav in the following terms:

“Regular inspection and cleaning of CTC strainers carried out, after the cooldown and loading operation at [the Bonny jetty]. Small amounts of debris, mainly consisting of mixture of granulated and very fine particles found in each strainer …. Debris tested for iron presence by the magnet and small amount of iron particles found. At the time of cooldown and slow loading at [the jetty], vessel still had 1008 m3 of heel on board, originating from Cameron LNG terminal. During the cooldown operations, pump performance was satisfactory and no pressure drops across the strainers were observed.”

The vessel sailed at 19.30 on 31 August 2011, and the cargo was discharged without incident at Tobata on 27 September 2011.

Pronav had appointed Inchcape Shipping Services (“Inchcape”) as agent for the discharge at Tobata, and on 29 August 2011 they had requested certificates in order to comply with Japanese customs. Pronav sent copies of four certificates provided in July 2009 by Nippon Kaija Kentei Kyokai (“NKKK”) about the CCMS: about the accuracy of the gauge for the cargo level, the cargo’s temperature, the pressure in the tanks and float gauges. In an email of 29 August 2011 Inchcape pointed out that the NKKK documents were surveyors’ reports, not customs verifications, but, more importantly for present purposes, the certificates were valid for two years in 2008 and that, although the period had been extended to three years, they would no longer cover discharge in September 2011. Apparently in the event this did not hamper discharge at Tobata, but this evidence shows that, at least as far as Tobata discharge was concerned, the vessel needed to renew these certificates.

It was in any case planned that the CCMS would be inspected and overhauled at Subic Bay. This was included in the standard dry-docking specification used by Pronav. Further, on 7 September 2011 Captain Baricevic wrote to Pronav that during loading at Bonny the float gauge in no 3 tank had jammed and parts needed repair or replacement and that the float gauge in no 4 tank was inaccurate. Pronav replied that it was planned that the float gauges would be overhauled and re-calibrated at the next dry-docking, and that they had arranged “specialized vendor (CIS [Singapore]) every drydocking period”. They promised to contact Cryogenic Instrumentation Services Pte Limited (“CIS”) about the problems that he reported.

On 28 September 2011 the “LNG Gemini” proceeded in ballast to KSS, arriving at Subic Bay on 4 October 2011 but berthing only on 24 October 2011. During the voyage, in order to prepare the tanks for the inspection, the vessel aerated them, and on 1 October 2011 the booster blower collapsed. Mr Milosic reported “abnormal vibrations”, “banging” inside the casing and “A big cloud of rust went out of the blower’s casing”. He said, “We can expect to find a significant quantity of fine dust in rust in the cargo [tank] spaces”.

The vessel went to the KSS for her scheduled regulatory inspection, the intention being to deal with the structural matters that had been identified on the ABS inspection in March 2011. She also was inspected in order for class to credit her with a dry-docking, either (as ABS had agreed subject to the agreement of the local surveyor) by actually dry-docking the vessel or by an underwater inspection or by examination of her rudder pintle bearing. Pronav had recommended dry-docking in order to treat her underwater hull and apply antifouling paint. Mr Cain initially hoped to avoid dry-docking, but in the event it proved to be necessary.

Pronav had recommended in a dry-docking specification for 2011 that planned repairs should not be done at KSS, because they were understood to have little experience of repairing LNG carriers, and recommended that “experienced Singapore shipyards” be used. AOM did not accept the recommendation: Mr Cain explained that the decision was driven by financial considerations, KSS’s price being some $1 million lower. It was suggested to Mr Cain that AOM did not expect to find Cameron debris in the tanks because otherwise the vessel would have been ordered to a shipyard with more experience of LNG carriers. I am not persuaded of this: there is no reason to reject Mr Cain’s explanation.

The vessel was at KSS until 14 December 2011. Mr Cain attended there for AOM, arriving on 2 October 2011 and staying until 31 October 2011, and then returning from 14 November 2011 to 11 December 2011. No arrangements had been made for work on the pumps. The arranged works included (i) planned steelwork, (ii) inspection of the cargo tanks for debris, (iii) removing, disassembling and cleaning the relief valves and (iv) work on the CCMS. With regard to the CCMS, Mr Kristic reported on 12 November 2011 that the specialists had started work “for NKKK certification”, but apparently the work was delayed and eventually testing was done on 29 and 30 November 2011. The transducers in the tanks were opened and, according to Mr Cain, a fine film of debris was found inside them. It was blown off and the transducers closed. Specialists from CIS came from Singapore to test that the CCMS was operating properly and correctly calibrated.

AOM always planned that the cargo tanks should be opened and inspected. The standard specification used by Pronav unsurprisingly shows that this is routine and ABS wanted at least three holds (nos 1,3 and 5) examined. The five tanks were inspected between 26 and 28 October 2011, and there was debris in them all. Captain Balog saw it, and in his statement he described tanks nos 1 and 2 as follows: “There was no sign of debris on the upper parts of the cargo tank, central pipe tower or access ladders. There was an accumulation of debris inside the pipe tower around the bases of the cargo pumps and the inside wall of the tower. There was also a significant amount of debris near the light holes and along the flat base of the tank”. He said that similar debris was found in the same places in the other tanks. However, according to the description in Pronav’s Cargo Tanks Inspection Report, which is more detailed, and to my mind more likely to be reliable, more debris was found by way of the tower pipes in tanks nos 1, 2 and 5 than in tank 3 (where there is a “small amount”) and in tank 4 (a “very small amount”).

Mr Cain said that “as we had feared, the crew found a significant quantity of debris in each of the tanks. The debris had settled on the tank bottoms and around the base of equipment in inside the central pipe tower”. He described it as being a similar colour to that found in the manifold filters at Cameron Terminal and the strainers on the voyage to Bonny, and not the fine white or grey powder that (according to Mr Cain and Mr Marković) is sometimes found in small quantities on the bottom of the cargo tanks or in strainers. However, I do not regard his observations as sufficiently specific to support AOM’s case that what was found in the tanks at Subic Bay was Cameron debris. Indeed, it became apparent in cross-examination that Mr Cain went into only one of the tanks, he believed into no 2 tank, which makes the less reliable his description of the debris in the tanks generally.

Pronav sought advice about the debris in the tanks from Atlas Copco, which Mr Cain described as a “cautious” approach. On 31 October 2011 Mr Cain asked Mr Vrenko what “Carter Pump” recommended and said that AOM needed “Something in writing”. Mr Vrenko replied that Altas Copco had recommended in July 2011 that they carry out “the highest level of pump overhaul” (“Pull out STATOR for cleaning and also lower end level assembly due to a pocket space”), and that Pronav would ask about a written recommendation. Apparently nothing had been received by 11 November 2011 because Mr Cain wrote again to Mr Vrenko, saying that, “Without this owners cannot bring a claim against Golar”. He asked that at least a representative from Atlas Copco visit the vessel and provide a report.

Atlas Copco sent what they called a “General Recommendation Letter” dated 11 November 2011 and signed by a Senior Operation Specialist, Mr Masateru Takata. The broken English is not easy to understand, but Mr Takata clearly misunderstood the position: he expressed “great shock” that there had been “iron rust” contamination, and advised that the “receivers” should refuse to have the cargo unloaded. The vessel should proceed, he said, to a shipyard “gas free” for works, including removing the main cargo pumps and the cool-down pumps for thorough cleaning. He also recommended that all LNG pipe lines should be cleaned with a vacuum cleaner, that cryogenic valves should be removed and cleaned, and that all check valves and expansion joints should be removed for cleaning. He explained that even a tiny accumulation of iron rust might conduct electricity. He said that, as manufacturer of the pumps, Atlas Copco would like to remove the iron rust as soon as possible, quoting a price of “18 million yen or so” on the basis of work at a Japanese yard and saying that it would be more expensive to have the work done elsewhere. Mr Vrenko commented on the recommendation in an email to Mr Cain of 11 November 2011: he said that Mr Takata was “going a bit far” in his proposition that there be “detailed cleaning” of the whole system, and said that this was because Pronav had provided “minimum of information”: “only the sediments’ nature was described”. And he said that, if all Mr Tokata’s recommendations were carried out, the work would cost some $2.5 million and take about three weeks.

Already before Atlas Copco’s letter, AOM had it in mind to have the pumps inspected in Singapore. On 7 November 2011 Mr Cain told Mr Vrenko that they needed to look at all the options, including Singapore, and Mr Vrenko replied that they would instruct Mr Nicola Kristic, a marine superintendent with Pronav, to “investigate [Singapore] option”. On 16 November 2011 Mr Cain told Mr Merrell that, after discussions with the Master and Mr Kristic, the preliminary plan was to open and inspect two main cargo pumps and one spray pump as a “representative sample”, but that the work could not be done at KSS. He said that the vessel had to go to Singapore, and that the work on the three pumps would take until 19 December 2011 and if work was required on all the pumps it would take until 30 December 2011. More would be known, he added, after Pronav had spoken to Atlas Copco again.

The suggestion that only sample pumps be inspected was apparently short-lived: on 17 November 2011 Mr Christansen was toldthat because Cameron debris had entered the tanks AOM intended to overhaul all the pumps. On 19 November 2011 Mr Cain reported to Mr Merrell again: after further discussions between Mr Kristic and Atlas Copco, the “strong” recommendation was that all the pumps (the ten cargo pumps and five spray pumps) be removed, dismantled and overhauled at a price of $520,000 (which would, he said, mean that AOM could claim from Golar $619,000, including the cost of fuel for the voyage from Subic Bay to Singapore). Atlas Copco provided quotations for the work on the pumps on 2 December 2011: ¥10,427,351 for labour and ¥7,970,000 for parts, mostly bearings. On the same day Pronav placed a formal order for the work and the parts.

On 14 December 2011 the Vessel proceeded from Subic Bay to Jurong, arriving on 18 December 2011. Mr Cain was already there. Mr Nygård and a Mr Idar Herland of Golar were at the shipyard for another project, and Mr Christiansen asked them to observe the work. An exchange between Mr Merrell and Mr Vrenko shows that they knew that Golar were unhappy about how much work was being undertaken: Mr Merrell wrote that Golar denied responsibility for it, and that the pumps should not be removed from the vessel. Mr Vrenko replied that the “scope of overhaul is exceptional, which means that the motor stators must be removed, high pressure steam jet cleaned, baked in the oven and installed back inside the pump housing”, and that the work could not be done with the pumps in the tanks.

On 19 December 2011 Mr Nygård asked Mr Cain whether they planned to open all ten cargo pumps and when they would be inspected. Mr Cain replied on 19 December 2011 that “Based on the debris found in the tanks and strainers and on the recommendation of pump manufacturer, all main (10 ea) and spray (5 ea) pumps require overhaul”. After being chased about when the inspection would be, he told Golar on 20 December 2011 that they planned to disconnect the pumps that day and to start lifting the pumps from the tanks on 21 December 2011.

The work of dismantling the pumps began on 21 December 2011 in a “clean room” at Jurong. It started before Mr Nygård arrived, but he observed it through the windows of the clean room, and he saw no sign of damage or unusual wear and tear to the pumps. Engineers from Braemar attended on 22 December 2011. By the time that they arrived, all fifteen pumps had been taken from the vessel to the workshop, and all but three cargo pumps (the portside pump from hold no 1, the starboard pump from hold no 3 and the portside pump from hold no 4) had been dismantled. They said in a report dated 5 January 2012 that there was no sign of mechanical damage, cracks or scoring marks on any of the cargo pumps, impellers or casings; that “grains of fine metal particles were found on the internal surfaces of the stator coils and bearing area” on the three pumps that were still being dismantled; and that they had been told that metal particles were similarly found in the other cargo pumps. Their report does not refer to the spray pumps (except that they had been transferred to the workshop before their engineers arrived).

Mr Christiansen arrived at Jurong on 23 December 2011, and by the end of that day all the cargo and spray pumps had been dismantled. On 23 December 2011 Mr Christansen said in an email to Mr David Ratliff of Golar that AOM’s concern about the pumps was a “storm in a glass of water”, and that “These pumps ‘eat’ pebbles”. Some forensic play was made of this, but he was simply using a figure of speech in an internal email to say that the pumps were too robust to be damaged by fine particles.

Mr Bo Johansson and Mr Jeff Poon of Viking arrived on 24 December 2011. In his witness statement Mr Cain said that Viking argued that only a sample of the pumps need be checked, but I reject that evidence: by the time that Viking’s representatives arrived all the pumps had been dismantled, and indeed two had already been returned to the vessel. Mr Cain also said in his statement that he “invited Viking Engineering to inspect the tanks but they declined”. In cross-examination he said that he had spoke to Mr Christiansen and not to Viking: it is not clear whether or when this invitation was given, but nothing turns upon it.

Mr Johannson examined the dismantled parts on 24 December 2011. He saw no damage and nothing that would prevent the pumps from working normally. He saw no debris except for some red rust on the electric motor rotors of the spray pumps, and Mr Takata told him that all the motor rotors had had rust that had been polished off: as Mr Johansson recorded in his inspection report of 28 December 2011, Mr Takata told him that “some of the electric rotors were very rusty before cleaning”. He also told Mr Johansson that it was intended to replace the ball bearings in view of their length of service. Mr Johansson examined all the bearings “one by one carefully” and he rolled the smaller bearings in his hand: he found nothing wrong with them. Mr Johannson is an experienced engineer and was an entirely reliable witness: I accept that the bearings that he examined were undamaged and infer that the other bearings were not damaged either.

Atlas Copco sent Pronav a report dated January 2012. The descriptions on the ten cargo pumps were similar. They were said to be in “good order”, apart from the stator coil, the conduit box and the upper and lower bearings. The stators and the conduit boxes were said to be “contaminated”, and this was because of the fine dust. The stator coil was cleaned and baked, and the conduit box was cleaned. The bearings were all replaced. The reports on the five spray pumps were also similar, and they too were reported to have been in “good order” apart from (i) the stators, which were again reported to be “contaminated” and were cleaned and baked; (ii) the upper and lower bearings, which were replaced; and (iii) the gaskets. It was not suggested that the condition of the gaskets is significant for present purposes.

Mr Cain’s evidence was that it was clear when they were dismantled that the pumps had not suffered “significant” damage. I conclude that they had not been damaged at all. There was, as Braemar, Viking and Atlas Copco reported, dust or powder in the pumps, but I conclude from the evidence that neither the amount nor the size of particles was particularly remarkable. Mr Cain described it as “very fine powder” around the bearings, and said that it “appeared mostly to consist of ferrous material”. Mr Nygard said that there was “just dust, nothing more than normal”.Mr Takata told Mr Nygard that the pumps were “a little dirty but that there was otherwise no problem with them”, and told Mr Johansson that there was “a bit of dust” in the pumps.

Why were the bearings replaced? Mr Cain said in his statement that, because “debris” was found in the pump bearings, Altas Copco advised that the pumps be cleaned and that new bearings be installed. In one answer in cross-examination he said that bearings “are always replaced when you open up a pump, particularly a close coupled pump like this where the fluid actually passes through the bearings”. But later he said that the bearings were replaced “because debris had been run through them” and that it was impossible to know whether they were damaged without dissecting them; and that otherwise the decision about replacing them would have depended on their running hours. New bearings had been ordered from Atlas Copco before the pumps were opened. Mr Takata told Mr Johansson that they were replaced “due to many years of service” and not, as he told Mr Johansson in reply to his specific question, because of debris or powder or dust in the pumps. I accept this. The bearings were not damaged and they were not replaced because of debris found in the pumps.

Clause 30 of the charterparty prohibits the shipment of “acids, explosives or cargoes injurious to the Vessel”. If the charterers ship prohibited cargo, any damage to the Vessel and any time taken to repair it are “for Charterers account”, and there was no dispute that therefore the Owners are entitled to be indemnified in respect thereof. However, this specific provision is without prejudice to other remedies that the owners might have for the breach of the charterparty.

AOM’s pleaded case is that “On examination [at Subic Bay] the ship’s cargo pumps and tanks were found to be contaminated with metallic particles that had been introduced into the cargo loaded at Cameron … causing abrasion, rust and risk of electrical short and pump failure”. In the end AOM did not argue that the Cameron debris had caused physical damage. No evidence of fact supports the claim that the particles caused “abrasion” or “rust” or any other damage to the pumps or tanks or anything else, and Mr Shortall accepted that, even if LNG loaded at the Cameron Terminal carried aboard small hard particles as AOM allege, they would not necessarily have caused damage.

Although clause 30 is in the terms of clause 28 of the standard Shelltime form, Counsel were not aware of any relevant authority about what “injurious to the vessel” means. Mr Saunders submitted that a cargo may be “injurious to the Vessel” within the meaning of the clause without causing any physical damage to her. First, he argued that a cargo would be injurious to a trading vessel (such as the “LNG Gemini”) if, after it was delivered, she needed to be cleaned and so taken out of service for the purpose: “the vessel in the claimants’ hands is an instrument of trade and anything that is harmful or tends to harm the vessel as an instrument of trade is harmful to the vessel”. He cited the Oxford English Dictionary (2^nd Ed, Vol VII), which gives “deleterious” as a meaning of “injurious”, and argued that clause 30 prohibits a shipment if it is deleterious to the use of a trading vessel for trade.

I reject that submission. Even if the word “injurious” might have so wide a meaning in another context, it is not its natural meaning in clause 30. Clause 30 is directed to physical damage: it expressly covers two types of cargo which might cause physical damage to the vessel, acid and explosives, and the inference is that it also covers other cargoes that also might cause physical damage. This interpretation is corroborated because the clause is particularly concerned with repairs of damage caused by such cargoes: “repairs” connotes physical damage. Further, clause 30 provides an indemnity for time lost to do repairs, but not for time lost by the vessel for other reasons, such as cleaning.

Mr Saunders’ alternative submission is that a cargo can be “injurious” to a vessel without actually causing damage to her if it is of a kind that has atendency or propensity to cause damage. I accept that submission: after all, a cargo of acid or a cargo of explosives might not in fact do damage. I did not understand Mr Eaton to argue otherwise.

AOM say that metallic particles in the LNG cargo created a risk that the vessel might be damaged in that:

There was some evidence about whether, after the cargo was loaded at the Cameron Terminal, AOM were concerned that debris might cause damage. This seems to me of peripheral significance: it does not bear upon whether the cargo was in fact injurious to the vessel (and neither Mr Saunders nor Mr Eaton suggested that it does), and has little relevance to what damages would be recoverable if AOM succeed on liability. Mr Cain’s evidence was that, after the problems experienced at the Cameron Terminal, AOM were concerned about whether debris had affected the vessel’s cargo equipment and sensitive monitoring systems. However, as Mr Eaton pointed out, this is not borne out by the contemporary documents. When Mr Merrell asked Pronav on 10 May 2011 whether there was any danger to the cargo pumps, Pronav’s response brushed the question aside almost causally: “Bearings never like sediments but our concern is the receiving terminal”. AOM did not contact Atlas Copco in May 2011, and when the vessel was surveyed after discharge at Senboko they did not express concern that debris might have damaged the pumps or the sensors. When it was reported in July 2011 that there was debris in the strainers, Pronav’s response of 29 July 2011 shows that they did not perceive an urgent problem and thought that it could await the next dry-docking. I do not accept that before the vessel went to Subic Bay AOM thought that there was any serious risk that debris had damaged the vessel.

In support of their case that the cargo was injurious to the vessel, AOM rely (i) on evidence about the loading operation at Cameron; (ii) on evidence about debris later found on the vessel during subsequent voyages; and (iii) on evidence about the state of the tanks and pumps at Subic Bay. I consider each of these, but first I observe that it is not unusual for an LNG cargo to include some small foreign particles. For example, Mr Magazinovic said that after the cargo was loaded at Bonny a “small amount of debris was found in the manifest strainers but no more than we would usually expect”. It seems to me that, while AOL might properly complain about the unusually large quantities of sediment discovered in the strainers at the Cameron Terminal, they cannot sensibly contend that there should not have been any particles at all in the cargo. Moreover, the surveyors who were asked to report on the problems during loading at the Cameron Terminal, LOC and Braemar, did not express any concern about the risks to the vessel such as AOl allege.

The debris found in the strainers at the Cameron Terminal was said by the witnesses from the ship to be “similar” (Captain Balog) and “exactly the same” (Mr Marković) on the two occasions that the loading operation was suspended. It was variously described as “sediment” (Mr Marković), “powder” (the Terminal’s log), “soil” (the vessel’s log) and “soil, dirt or sand” (the letter of protest). Captain Balog described “lines of brown debris” in the manifold that continued up to the strainer, Mr Marcović said that there was a “large amount of red-brown material”, which was “a bit like soil”, and Mr Magazinovic spoke of “deposits of reddish-brown sediment running along the bottom of the spool piece and manifold pipe”.

Samples of the debris from both stoppages were analysed by SGS North America Inc (“SGS”) for Golar and Sempra. The analyses showed iron in different amounts: a sample from the first stoppage contained 9.9% iron (and also 8.5% silicon), and one from the second contained 32.3% iron (and 5.6% silicon). Sempra had tests conducted by Structural Integrity Associates Inc by way of an energy dispersive x-ray spectroscopy analysis, and different samples showed iron content of 33.4%, 34.4% and 31.2%. The results might suggest that their samples were from debris collected after the second stoppage, but there is no convincing evidence about this.

The strainers did not catch only particles that were too large to pass through the fine mesh: some smaller particles too were caught because the mesh had become clogged with larger particles. (The mesh was described by Captain Balog in an email to Pronav of 9 May 2011 as being “completely clogged with soil”.) This is shown by an analysis of the size of the particles by Particle Technology Labs (“PTL”). They analysed three samples: it is not clear from where they were taken, but one was described as “LNG 2 & 3 (Manifold)” and so was presumably from a manifold. They show that between 78% and 84% of the particles (in volume) were themselves small enough to have passed through the filter. Moreover:

Cameron debris might have entered the tanks in two ways. Particles smaller than 0.149 mm (or 149 microns) could have passed through the fine mesh of the strainers. Mr Clucas said that some will have done so, and Golar accept this. Further, as I have said, the by-pass valve on arm no 39 was opened at 21.35 on 9 May 2011 in order to help to drain the arm. Mr Shortall said that this “potentially allow[ed] heavily fouled liquid cargo to enter the tanks”. This LNG by-passed the filter and so it might have carried larger particles aboard. It is impossible to measure how much LNG by-passed the filters, but I consider it unlikely that a significant amount of debris was carried into the tanks in this way. LNG is not a dense cargo, and the pumping operation had been stopped some 30 minutes or more before the bypass valve was opened. During that time any debris, particularly iron debris, will have settled at the bottom of the pipe. (Mr Shortall, said that the density of LNG is about 450 kg per cubic metre, and that of iron is about 7,880 kg per cu metre.) Further, as Mr Clucas observed, the photographic evidence does not appear to bear out Mr Shortall’s description of “heavily fouled” liquid.

Mr Shortall suggested another reason to infer that a significant amount of debris was carried into the tanks before the filters were clogged. He surmised that smaller and lighter particles would have been carried along the pipe before larger, heavier particles that caused the blockage. I am not persuaded of this: it assumes that the large particles were all heavy, but this overlooks that they probably comprised a wide variety of materials. I see more force in Mr Clucas’ observation that the filters had clogged although the LNG had been loaded slowly: he calculated a rate of about 300 cu m per hour whereas the shore lines had probably been designed for rates of about 5,000 cu m per hour. The first blockage occurred shortly after the loading valves were opened from 40% to 100% capacity. This suggests, as Mr Clucas argued, that the debris that caused the first blockage was disturbed in the pipe when the loading rate was increased.

Mr Clucas made a brave attempt to estimate how much debris might have entered the tanks on what he called a “worst case scenario”. He did so by supposing, from his understanding of the PTL report, that a third of particles were stopped by the filters and twice as many passed through them, and by estimating the amount of deposit in the filters. He thought that “the worst case scenario would be that about 20 litres in total passed the filters and into the Vessel’s cargo systems”. This, he observed, if spread evenly between the pipe columns would have resulted in a covering about 0.5 mm deep, which he considered would be “practically invisible” and inconsistent with the description of debris in the tanks at Subic Bay. Mr Shortall disputed this reasoning, not least because, as Mr Clucas accepted in cross-examination, the PTL report does not in fact support the supposition that the filters captured a third of the particles. Mr Shortall thought that any attempt to calculate the amount of debris was “no more than a guess”, but did put forward a suggested method of doing so, which, after correcting admitted errors, would indicate that some 21 to 28 litres entered the tanks. Although this is strikingly close to Mr Clucas’ 20 litres, I do not consider these assessments sufficiently reliable to assist. Mr Clucas agreed in cross-examination that his calculation gave only an “order of magnitude”, and indicated only that there was unlikely to be more than 200 litres or less than 2 litres of debris.

I consider, therefore, that the evidence about the loading justifies only the inference that some small particles passed through the filters before they clogged on the two occasions and that some of the particles, maybe as many as a third of them, were metallic. However, in my judgment no more can be inferred. It does not establish that larger particles were loaded in any significant number, and it provides no basis for inferring how much particulate debris was loaded.

The evidence about the debris found in filters on the voyages to Senboko, Singapore and Bonny does not advance AOM’s case. The debris found on 14 May 2011 suggests that after the second blockage some more particles, including presumably metallic particles, were carried through the filters. But again they will have been smaller than 0.149 mm. This is corroborated by the problem-free discharge at Senboko, and consistent with the finer 150 mesh filters in the spray pump catching particles of less than 0.149 mm that had passed through the 100 mesh filters in the manifolds.

What of the debris in the tanks and cargo systems when the vessel was at Subic Bay? AOM asked me to infer that the debris found around the base of the pipe towers and pumps was carried on board with the LNG loaded at the Cameron Terminal. This is not shown by any analysis of the debris or any evidence about its appearance: as I have explained, I attach no significance to Mr Cain’s evidence that the debris in the tanks had a similar appearance to the debris in the filters during loading, and there is no credible evidence of analysis of the debris in the tanks. (Mr Cain said in one answer in cross-examination that he believed that a test was conducted for “13 or 14 elements” and “they were all consistent with the same material, the nonferrous material that was found in the tanks – or ferrous material found in the tanks”. But his evidence was too vague to be reliable.)The real thrust of AOM’s argument is that relatively soon after the unusual problems experienced at the Cameron Terminal an unusual quantity of debris in the tanks was found in the tanks at Subic Bay, and, in the absence of any other explanation, it is to be inferred that that the bulk of it was Cameron debris.

The problems experienced during the loading were indeed unusual. Loading of an LNG cargo is seldom suspended because a manifold strainer has become clogged, still less suspended twice. Mr Eaton properly observed that the filters on the “Gemini” and her sisterships were unusually fine, and otherwise the strainers might not have been blocked, but what happened at the Cameron Terminal was unusual even for the Aquarius vessels with their fine filters.

What of the debris found at Subic Bay? Both Captain Balog and Mr Magazinovic said that they saw “a significant quantity in the tanks”. Mr Cain echoed this, saying that the crew found “a significant quantity of debris in each of the tanks”. They all used exactly the same expression in their statements, which might suggest that their assessments were not independent of each other, but I accept that they found more debris than is usual. This conclusion finds some support in photographic evidence, although I found the photographs difficult to interpret.

However, I am not persuaded that the evidence supports AOM’s attractively simple reasoning that the bulk of the debris found in the tanks was Cameron debris, or that it establishes AOM’s case that Cameron debris created a risk of damage to the vessel. First, Mr Clucas considered that the amounts of debris described as being in the tanks was much more than would have been loaded at the Cameron Terminal. Admittedly his reasoning depended to some extent on his questionable calculation that there might have been some 20 litres of Cameron debris, but I accept that the very quantity described indicates that there was at least one other major source. This inference is reinforced because, as Mr Shortall and Mr Clucas agreed in their joint memorandum, some Cameron debris will have been removed from the tanks on discharge at Senboko, and more was caught in the filters of the spray pumps during the voyage to Bonny. More would have been carried from the tanks when the subsequent cargo was discharged at Tobata. I cannot assess how much was removed, but the total would, I think, have been significant.

This leads to the question whether there are other possible explanations for the debris found at Subic Bay. It is agreed that some of the debris in the tanks was not Cameron debris. It included cable ties in all five tanks: it is not clear how or when they entered the tanks: Mr Shortall might well be right that they would not have survived the cold of a cargo voyage and were introduced at KSS, but the fact remains that not all the debris that Mr Cain and others saw was Cameron debris. In tank no 3 there was also what was described as paper-like material, which it is agreed could not have come from the Cameron Terminal.

There are other possible explanations for particles, including metallic particles, in the tanks.

Golar had another argument that the debris found in the tanks was not Cameron debris. Debris was found in all the tanks. The evidence does not show how evenly it was distributed between them. Mr Clucas over-stated the point when he said in his first report that all the tanks “seem to have a similar amount of debris”, but it suffices for Golar’s argument that significant amounts were found in all the tanks. Mr Clucas explained that, because of the design of the tanks and loading methods, during loading LNG would not have flowed into them all simultaneously. If, as seems likely, the Cameron debris mostly was loaded in a relatively short period, the LNG carrying it will not have run into all the tanks, but taken the easiest route to one or possibly two tanks. The distribution of debris at Subic Bay does not reflect this. It is a pointer that it was not predominantly Cameron debris, and in my judgment a convincing pointer.

I accept that some of the debris found in the tanks at KSS was Cameron debris, but in my judgment AOM have not proved that much of it was shipped by Golar at the Cameron Terminal. I consider it more likely that the greater part of it was from elsewhere. In any event, I am not persuaded that Golar shipped a cargo injurious to the vessel or that particles in the LNG created potential dangers as AOM contend.

Although Mr Shortall identified potential problems that might theoretically have been caused by particles, apparently the industry generally does not share his views. SIGTTO introduced their recommendation that strainers be fitted in the manifolds of LNG carriers after incidents in which LNG carrying solid matter was discharged into receiving terminals and in response to concerns expressed by terminal operators or the authorities supervising them. It was not introduced in order to protect vessels from damage from particles. Moreover, as I have said, SIGTTO recommend 60 mesh or 20 mesh filters, which prevent only larger particles from entering the cargo system, and these are usually what LNG carriers use. Mr Shortall accepted that “the industry does not contemplate that particles of up to 0.25mm are going to cause any damage to the LNG system”. There is no cogent reason to think the industry’s standards and practices too lax or that those responsible for them are wrong not to disregard particles small enough to pass through 60 mesh or 20 mesh filters. Cases of damage from LNG cargo are rare: Mr Shortall said that he knew of no reported cases in which debris had caused damage to pumps. Mr Clucas, with his long experience in the industry, said that he knew of only three or four reported cases of damage of any kind from debris entering into cargo tanks. I accept that that evidence gives the true picture, and there was no suggestion that the “LNG Gemini” had particular features that made her particularly vulnerable to damage from particulates.

With regard to the bearings, Mr Clucas said that Carter pumps are designed to operate with small particles in the LNG that passes through them, and this is corroborated by the fact that, as Mr Clucas explained in unchallenged evidence, LPG carriers have the same type of Carter pump, LPG being a much dirtier cargo than LNG. The design of Carter pumps on LNG carriers pre-dates the SIGTTO guidance, and there is no evidence that particles damaged pumps before strainers were commonly used. Nor does Atlas Copco’s documentation about the pumps indicate that this is a danger: a document called “Recommendations for protective devices” provides, as the title suggests, advice to operators about avoiding damage to pumps. It advises about over-current protection and about not running the pumps dry, but nothing is said about potential damage from particles or dirty cargo. Mr Clucas agreed that “gross contamination” from hard abrasive metallic particles could damage bearings because the bearings are lubricated by the LNG, but made clear that he was referring to a degree of contamination far in excess of anything that could on any view have resulted from the Cameron debris. I add that even Atlas Copco’s General Recommendation Letter does not suggest that iron rust might damage the bearings.

As for the electrics, Mr Shortall considered that metallic particles could become trapped in the electric motors on the cargo pumps and so could cause short circuits and a motor failing. Mr Shortall’s expressed concerns were not based upon any experience of such problems in submerged pumps. I was impressed by Mr Clucas’ response, which was based on his experience. He explained that the pumps are coated to protect them against short circuiting, and that he had never known metallic particles cause short-circuiting either in LNG or LPG carriers, notwithstanding:

The level sensors were a capacitance type probe; that is to say, the capacitance changes when the level of the LNG rises between the prongs, and the reading is converted electronically to indicate the level of LNG in the tank. It was suggested that contamination would affect the signal between the probes and so the level reading. Again, in my judgment this was speculation and not based on any actual experience of problems of this kind. It is unrealistic, to my mind, to think that equipment in the cargo system is designed to operate without any metallic contamination. Mr Clucas’ experience shows that it is more robust than that.

I therefore conclude that AOM have not proved that Golar were in breach of the charterparty.

Despite this conclusion, I shall consider whether AOM would be entitled to damages or other relief if I am wrong and on the basis that the LNG cargo contained particles of a kind and size and sufficient in number for the cargo to create a risk of damage to the vessel. Their claim includes extra expenses incurred when the vessel loaded at Cameron and costs associated with work done by KSS and at Jurong. On AOM’s case, and according to Mr Cain’s evidence, this work and expenditure was a reasonable response to the problems during loading at the Cameron Terminal and their reasonable concerns about the vessel’s pumps and the cargo system generally. The claim for $1,933,993 is pleaded under ten heads as follows:

The “crew overtime” claim arises from a proposal by Pronav that the crew be paid a bonus. In a letter to Captain Balog dated 17 May 2011 they said that the crew completed “a challenging task ... in an excellent manner”, and that the Owners had approved “compensation for performing substantial extra work and overtime” in the sum of $8,000 by way of a “one-off payment due to the lengthy nature of this unusual loading process”. I reject this claim. First, the clogged strainers were not the only problems in the “unusual loading process”. There were also problems during the gassing-up process and because GNG could not be returned to the Terminal. Mr Cain said in his third statement only that the bonus was paid because of “the exceptional difficulties experienced during the loading of cargo”, and there is no evidence whether it was paid because the strainers blocked or because of the other problems. Secondly, the clogged strainers and any additional work associated with them were not the result of Golar shipping dirty LNG. On the contrary, they were the result of the strainers preventing particulate matter from being shipped.

Next, the claim for $17,023 which is pleaded as being for “Tank inspection/cleaning and valve inspection”. According to Mr Cain, this represents sums paid to KSS for four invoiced items: (i) “Safety valve – cargo tank”, for $5,003; (ii) “Temporary protection shelter/cover in 5 cargo tanks”, for $4,500; (iii) “Smaller valves in way of cargo tank domes”, for $5,400; and (iv) “Staging/lighting in way of work in cryogenic tech”, for $2,120.

The payment of $5,003 is for overhauling the cargo tank relief valves. AOM’s case is that Mr Cain arranged that the safety valves be disassembled and cleaned because of the incident on 13 May 2011, when he understood the safety relief valves were almost forced open. He explained that he was concerned that debris from the tanks had become lodged in the valve mechanisms, and therefore they might not open in an emergency or, if opened, they might not close. As I have explained, I do not accept that there was any reason to think that the valves were forced open or nearly forced open. In any case Mr Clucas pointed out that the layout of the Moss tanks makes it unlikely that the valves were contaminated: the inlets of the safety valves are comfortably above the maximum cargo levels permitted in the tanks, and moreover any debris would have been relatively dense and so at the bottom of the tanks. Mr Shortall suggested that vapours of gas generated during loading might have carried light particles to the top of the tanks, but this seems to me fanciful.

This is enough to reject the claim, but it faces another difficulty. The safety valves would in any case have been pressure tested on the bench for the purposes of the ABS survey. The invoice supporting the claim for $5,003 described the work as removal of the valves, sending them to the workshop, pressure testing, and refitting them on board. All that work, as it would appear from the standard specification, would have been done in any event. Mr Cain said that AOM decided to open the valves for cleaning only because of what happened at the Cameron Terminal: “We would have just pop tested them. There was no other reason to open them up other than because of the incident in Cameron”. He invited the inference, I think, that therefore the work was more expensive. However, the documents do not reflect this and AOM have not attempted to identify the amount of the additional cost.

In his statement Mr Cain said that the item for $5,400 was also for “assisting with the overhaul of the cargo tank relief valves”, but in cross-examination it was clear that he had no idea to what this charge related. There was insufficient evidence to prove this part of the claim, and moreover it probably faces similar difficulties to the claim for $5,003, the other part of the claim relating to the safety valves.

The charges of $4,500 and $2,120 were incurred because the tanks were opened. They were for temporary staging and weather protection, and in order to make it safe to enter into the tanks and work there. However, it was always intended that when the vessel went to KSS the tanks would be opened. Mr Cain said that because of the debris the tanks were open for longer (and so the staging and lighting required for longer), but no claim was presented for the additional costs, and AOM have not proved what they were.

The claim for $119,891 is also for work done at KSS. It represents a payment to CIS for providing technicians to inspect, overhaul and re-calibrate the CCMS. Their invoice covers mechanical works and electrical instrumentation works over a period of 24 days (together with incidental expenses, such as travel and accommodation). Mr Cain accepted that some mechanical work was done to repair the gauge floats and some electrical works were required to meet NKKK’s requirements. AOM have produced no evidence that enables me to identify the cost of additional work required because of the Cameron debris, and they have not proved this claim. Mr Cain said in cross-examination that, “If the system is operating correctly the testing of the system and reissuing the certificate is a one day or two day event”, and Mr Saunders invited me to rely on this to make a “pro rata” reduction in the claim. I regard Mr Cain’s evidence as too vague and unreliable to do so.

I come to the claim for $310,724 for the cost of fuel for the passage from Subic Bay to Singapore. Golar disputed that the decision to have the pumps overhauled by Jurong was necessary or reasonable. I must consider this on the basis that the cargo shipped to Senboko was injurious to the vessel because of Cameron debris for the reasons that AOM allege, including the risk that metallic particles might have damage to the pumps. On this basis, it was certainly reasonable of AOM to have the pumps opened, inspected and cleaned. The question is whether it was reasonable to have the work done in Singapore rather than at Subic Bay. If it was, fuel costs would be incurred and Golar do not dispute that they are properly calculated at $310,724.

The contemporary documents refer to four reasons that KSS should not carry out the work done, and they are stated as follows in an email from Pronav to AOM dated 28 October 2011:

“KSS has no “clean room” for this type of maintenance.

KSS has no experience in handling of cargo pumps between ship and shore.

There is no specialized/dedicated team on shipyard side for this type of work.

KSS has no logistic required to maintain the controlled atmosphere in the cargo tanks. High humidity in the cargo tanks may further affect the condition of pumps/bearings/in-tank electronics”.

AOM did not rely on two of these reasons to justify the decision to go to Singapore: they did not contend that experience was required to move the pumps from ship to shore or rely on the concern about a “controlled atmosphere”. Their arguments were that KSS did not have suitable facilities for the work and that they did not have experience of such work.

Mr Shortall considered that, because work on LNG carriers is “of a highly specialised nature”, it is usually done by a yard with previous experience. Mr Clucas agreed that submerged pumps of this kind have to be treated as precision equipment, but disagreed about whether the work could properly have been done at KSS. I prefer Mr Clucas’ evidence. It is clear from the instruction manual for the Carter pumps that it does not require either specialist tools or specialised expertise to disassemble them. It contemplates the owners working on the pumps annually, and nothing suggests that it requires specialist skills such as Mr Shortall suggests. In any event the work at Jurong was supervised by Atlas Copco, and the work could similarly have been supervised at KSS. Even if KSS did not themselves have employees who could do it, third party contractors could have been used.

Mr Shortall also questioned whether KSS had the facilities for the work. Atlas Copco’s maintenance manual suggested that pumps be disassembled in “a clean, dry, well lighted area”. Mr Shortall said that such areas will not be available in many shipyards used for general mechanical work, and their workshops were likely to be too dirty for disassembling pumps: the internal working of the pumps would be vulnerable to contamination. His view was that “no prudent owner would choose to carry out repairs to precision equipment such as submerged pumps, at [KSS] as the yard is not a suitable location for such work”. However, in my judgment, no credible evidence supports Mr Shortall’s assertion that KSS could not have found suitable facilities. Mr Clucas pointed out that, while some yards claim to have a “clean room” for overhauling sensitive cargo equipment, this refers only to a facility protected from the dust and dirt generally associated with dry-docks, and said that, if KSS did not have such facilities within the shipyard itself, they can usually be found nearby without difficulty. In his memorandum of 11 January 2012 Mr Cain explained the decision to have the work done at Jurong only on the basis that KSS “had no previous experience with working on cryogenic systems”, and did not refer to inadequate facilities. I conclude that AOM simply have not adduced evidence to justify on this basis the decision to have the work done at Jurong.

AOM have an alternative argument to justify the claims relating to the vessel going to Singapore: that they acted in accordance with the recommendations of Pronav and Atlas Copco, that it was reasonable of them to do so and that therefore the resulting costs are recoverable. According to Mr Cain, the decision that the pumps should be opened at Jurong, and not at Subic Bay, was “mostly driven by Pronav”. I reject that argument. The costs were not caused by Golar’s (supposed) breach of contract simply because AOM responded to it in accordance with the advice of others. But in any case I do not accept that it was reasonable for AOM to incur expenses in reliance on what Pronav and Atlas Copco advised. It was obvious from the General Letter of Recommendation that Atlas Copco completely misunderstood the problem about which their advice was sought, and there is no sensible explanation for this or why Pronav did not seek their advice on the basis of the true facts. Apparently AOM were not curious about this. I do not accept that in these circumstances they can justify their decision to have the work done by Jurong on the basis of the advice of Pronav and Atlas Copco.

I would therefore reject the claim for fuel costs. Golar had other arguments about this part of the claim, but if I am right that AOM have not justified the decision to go to Singapore, they do not need to rely on them. I shall deal with them briefly:

I come to the claim for time lost by the vessel. If I am right that AOM have not shown that the fuel costs are recoverable because they have not justified the decision to have the pumps overhauled in Singapore, I would reject this claim for similar reasons. If the fuel costs for going to Singapore are justified, in principle compensation for time lost would be recoverable. However, Golar contended that in any event the claim should be rejected.

The claim for $961,581 is based upon AOM’s contentions that (i) the work reasonably took from 06.24 on 14 December 2011 to 18.00 on 31 December 2011, a total of 17.4833 days, and (ii) the value of the time was $55,000 per day because that was the market rate for hiring the “LNG Gemini” in December 2011. If the claim is justified in principle, I would accept AOM’s claim for 17.4833 days’ compensation. Mr Eaton did not argue that the time taken to overhaul and clean all the pumps was excessive. Golar sought to contend that AOM could and should initially have opened only some of them, but this is easily answered. Even if it was unreasonable to open all the pumps (and I do not consider, once the cargo is supposed to have been injurious because the pumps might be damaged, that it was), no time would have been saved by testing a sample because, when it was inspected, debris would have been found, and therefore all the others would have been opened and cleaned in due course.

Golar’s real challenge to the claim for $961,581 was that no evidence supports a daily rate of $55,000, given that I refused AOM permission to rely upon this part of Mr Cain’s evidence. The only evidence of a market rate is that in the charter that AOM entered into with Golar the hire was $28,000 per day. In these circumstances AOM argued that I should defer the quantification of this head of damage and direct that it be assessed at a separate hearing. Golar argued that AOM have simply failed to adduce any evidence of the appropriate rate and therefore this claim should be dismissed. In my judgment, it would be wrong to defer assessment of this claim and to allow AOM a separate hearing to prove their loss. AOM had no real explanation for not serving their evidence in proper time and in accordance with the court’s directions, and there would be no justification for devoting further resources to rescue them from the consequences. The question, therefore, would be whether to assess this claim on the basis of a daily rate of $28,000, or to reject it because AOM have not proved their loss. Of course, the market would have moved between January 2011 and December 2011, and therefore the rate in the charterparty of 3 January 2011 is unlikely exactly to measure the loss of time when the vessel was at the Jurong. But I would conclude that, in the absence of better evidence, the charterparty rate should be used to quantify AOM’s claim for loss of use. This would result in an award under this head of $489,532 if the claim for loss of time had succeeded in principle.

Next, the charges made by the Jurong and Atlas Copco in respect of their work on the pumps. Jurong charged S$253,946 for their work on the cargo system, including incidental expenses such as moorage and wharfage. Atlas Copco charged a total of ¥18,387,351 for their work on the pumps in Singapore in accordance with their quotations of 2 December 2011. I accept Mr Cain’s evidence that he converted these sums into US dollars at the prevailing exchange rates. Golar did not argue that AOM should not have used the services of Atlas Copco, and there is no reason to think that Atlas Copco would have charged a significantly different amount had they been supplied in Subic Bay rather than at Jurong. Had AOM succeeded on liability, I would have upheld the claim for $238,398. If the work had been done in Subic Bay, although AOM would not have incurred Jurong’s charges of $196,446, they would have incurred corresponding charges. Had AOM succeeded on liability, I should have invited further submissions about whether and how I might recognise this.

The claim for “third party inspections” is in respect of fees charged by Braemar. When Mr Cain put AOM’s case together, he did not have the documents to support this claim, and included $4,000 as what he called “a relatively low estimate”; and this is reflected in AOM’s pleading. Although the estimated sum was in fact too low, Mr Saunders made clear that AOM limit the claim to the pleaded figure. They rely on two invoices of Braemar: one for $5,044 dated 28 September 2011 was in respect of an inspection at the Cameron Terminal; and the other for $3,406 dated 27 April 2012 was in respect of an inspection at Jurong. Both were addressed to General Dynamics’ underwriters, who instructed Braemer, and there is no evidence that AOM, or even General Dynamics, bore this expenditure. AOM have not proved this head of claim. I add that the fees for the first inspection were not apparently incurred because Golar shipped a dirty cargo, but because the strainers did their job and prevented particles being shipped.

I come to the claims in respect of “superintendent”: for travel in the sum of $23,126, and for expenses and “inspection” in the sum of $54,804. According to Mr Cain’s statement, the sum of $23,126 related to the cost of his own attendances at KSS and Jurong. However, in cross-examination he said that AOM intended to claim only for his travel to Jurong because he was in any event travelling to the Philippines on another matter. The claim for $23,126 was therefore not supported by cogent evidence. Although no doubt AOM incurred costs because Mr Cain travelled to Jurong to supervise the works, they have not proved what they were. The claim for them would therefore have to be dismissed.

The $54,804 relates to Mr Kristic attending at KSS between 26 and 28 October 2011 when the tanks were opened and between 12 November 2011 and 6 December 2011 when the CCMS and the safety relief valves were overhauled, and at Jurong between 16 and 31 December 2011. Mr Cain acknowledged that the claim for $54,804 was overstated and said that it should be for $52,791, the amount which AOM paid Pronav for Mr Kristic’s services. But AOM have not established that the reduced claim is justified:

Mr Eaton advanced another argument against some of the heads of claim, including the claims in respect of crew overtime and works at KSS. He argued that AOM have not proved that they incurred these costs and can recover in respect of them. Mr Cain thought that AOM had “ultimately funded” the bonus for crew overtime, and that payment for the KSS works was made by Pronav and then “we wire transfer the funds back to Pronav, I believe”. But he had no direct knowledge of this. Mr Eaton pointed out, with some justification, that Mr Cain was not careful in distinguishing between AOM, Patriot and General Dynamic and submitted that AOM have not proved their loss. This is a commercially unattractive argument, but that is not a proper basis to reject it. However, exiguous though AOM’s evidence was, I would have concluded that these losses were probably borne by AOM in view of Mr Cain’s understanding and the inherent probability that expenses of this kind would be met by the managing owners.

Finally, I should refer to the claim for an indemnity in respect of “any and all claims against [AOM] for contamination of cargoes carried on the ship caused by [Golar’s] breach of contract [as alleged in the particulars of claim] in loading contaminated cargo at Cameron”. There is no evidence that AOM have had any such claim made or threatened against them, and I cannot foresee what claims might be covered by any indemnity. In these circumstances, I would not grant a declaration of an entitlement to be indemnified: as Somervell LJ explained in Trans Trust SPRL v Danubian Trading Co, [1952] QB 297, 303, such a declaration will not be made in general terms if it risks precluding proper argument when a specific claim is advanced. Nor to my mind is this a case in which the court should postpone the determination of future possible heads of damage. Although McGregor on Damages (18^th Ed, 2009) rightly recognises at para 8-025 that the courts are generally more prepared in appropriate cases to defer adjudication of issues about damages, they only do so when there is reason to think that there might be a future claim. I would therefore in any event have refused this relief.

AOM have not proved that Golar shipped a cargo injurious to the vessel, and I therefore dismiss the claim. They have, in any event, failed to prove much of their damages claim. Golar are entitled to judgment on the counterclaim.