Sizer Metals Pte Ltd v Chubb Insurance Singapore Ltd [2022] SGHC 51

Introduction

The plaintiff, Sizer Metals Pte Ltd, ordered nine shipments of tin concentrate (also referred to as cassiterite) from Excellent Mining Company Ltd (“Excellent Mining”), a company incorporated in Rwanda. The transportation of the first five shipments of tin concentrate was uneventful and these drums of tin concentrate arrived in Penang safely. However, for the sixth to the ninth shipments of tin concentrate, on their arrival in Penang, the contents of the drums were found to have been swapped with iron oxide. The plaintiff had bought insurance coverage for all the shipments from the defendant, Chubb Insurance Singapore Ltd.

The present dispute concerns the sole issue of whether the thefts of the plaintiff’s tin concentrate in the sixth to the ninth shipments were covered by the insurance policy provided by the defendant. It is undisputed that the thieves swapped the tin concentrate loaded in metal drums with iron oxide for the sixth to the ninth shipments. The plaintiff claims the sum of US$1,154,508.94 or damages to be assessed under the insurance coverage provided by the defendant.

Background to the dispute

I shall set out the background facts which are largely uncontroversial.

The plaintiff is a company incorporated in Singapore and carries on the business of trading base metals. The defendant is an insurance company incorporated in Singapore.

On 16 September 2013, the plaintiff and the defendant entered into Marine Cargo Insurance Policy No 92359646 (the “Policy”) under which the defendant would indemnify the plaintiff against any loss, damage or expense arising out of the transit of tin concentrate from Kigali, Rwanda to the port at Dar es Salaam, Tanzania and thereafter to Penang, Malaysia.

The salient terms of the Policy were cll 8.1 and 11.1, which I set out below:

These clauses provide that the plaintiff’s tin concentrate would be insured only for the period after the “commencement of the transit” (the “Transit Period”).

On 15 September 2017 and 30 May 2018, the plaintiff entered into two sale and purchase contracts with Excellent Mining for the purchase of nine shipments of tin concentrate. The contracts referred to the tin concentrate as cassiterite.

I pause to set out the various stages involved in the transporting of the tin concentrate from Excellent Mining to the consignee, Malaysian Smelting Corporation Ltd (“MSC”) in Penang. This supply chain is undisputed but important to the present case:

(a) First, approximately 20 to 25 metric tonnes of tin concentrate for each of the affected shipments were procured by Excellent Mining from mines in Rwanda (including its own) and transported to Excellent Mining’s premises in plastic bags which were secured with Precintia tags. These bags were then emptied onto the ground and the tin concentrate was mixed. Samples of tin concentrate were collected for analysis by a representative of Alex Stewart International Rwanda Ltd (“Alex Stewart”). After samples were collected, the tin concentrate was weighed and poured into empty drums through a hole using a funnel in the presence of the Alex Stewart representative. Once the drums were filled up, the holes of every drum (comprising a bung hole and a ventilation hole) were welded shut and sealed with Precintia clips by the Alex Stewart representative, in the presence of a representative from the International Tin Supply Chain Initiative (“ITSCI”) and a Mineral Field Officer (“MFO”) from the Rwanda Mines, Petroleum and Gas Board (“RMB”). The lids of all drums were then applied with white alkyd paint by Excellent Mining’s personnel. The ITSCI shipment number and/or lot number and the addresses of both Excellent Mining and the plaintiff were written on the paint coating. As part of standard procedure, the Alex Stewart representative would take photographs of the sampling, weighing and packing process for record purposes. The sixth to the ninth shipments were sampled and sealed on 17 May 2018, 5 June 2018, 19 June 2018 and 28 June 2018 respectively. While waiting for a report on the samples by Alex Stewart, the sealed drums with the contents were kept in a yard surrounded by a compound wall in Excellent Mining’s premises in Kigali, Rwanda. The plaintiff asserts that the compound of Excellent Mining was protected by security guards 24/7 and closed-circuit television (“CCTV”) surveillance cameras.

(b) Second, for each shipment, when the drums were ready to proceed to Bolloré Logistics Rwanda Ltd’s (“Bolloré Logistics”) bonded warehouse (the “Bonded Warehouse”) in Kigali, Rwanda, they were loaded into a 40ft container in the presence of the Alex Stewart representative. The loading into the container was done outside the Excellent Mining compound as the compound was not big enough for the 40ft container. When the drums were loaded into the 40ft container the Alex Stewart representative then affixed temporary seals on the container’s doors.

(c) Third, the 40ft container was driven from Excellent Mining’s premises to the Bonded Warehouse. This journey took approximately 45 minutes to one hour.

(d) Fourth, at the Bonded Warehouse, the 40ft container containing the drums will clear customs. The temporary seals on the doors of the 40ft container were broken and a further inspection was conducted by representatives from Alex Stewart, Bolloré Logistics and the Rwanda Revenue Authority (“RRA”). After the inspection, the doors of the 40ft container were closed and sealed by the same representatives.

(e) Fifth, after clearing customs, the 40ft container was transported on land from the Bonded Warehouse in Kigali, Rwanda to the port in Dar es Salaam, Tanzania. The journey was 1400km to 1500km long and required overnight stops over several days.

(f) Sixth, the 40ft container arrived at Dar es Salaam and was parked at the Inland Container Depot.

(g) Seventh, in the presence of the representatives of the Tanzania Revenue Authority and Bureau Veritas (Bolloré Logistics’ surveyors) the seals of the 40ft container were broken and the drums were inspected by them. Thereafter, the drums were loaded into a 20ft container. The doors of the 20ft container were then sealed in the presence of representatives of the Tanzania Revenue Authority, Bureau Veritas and the liner shipping company.

(h) Eighth, the 20ft container was loaded on board a ship for a sea journey from Dar es Salaam to a port in Penang, Malaysia.

(i) Ninth, upon arrival in Penang, the seals of the 20ft container were broken in the presence of a receiver who would take delivery of the drums.

Under the first sale and purchase contract dated 15 September 2017, the plaintiff was to take delivery of the first six shipments of the tin concentrate at the Excellent Mining premises. This was the point of the “commencement of the transit” within cl 8.1 of the Policy (see [6] above). Thus, it is undisputed that the insurance coverage for the sixth shipment of the 27 drums of tin concentrate weighing 18 metric tonnes started when the tin concentrate left Excellent Mining’s premises, ie, from the third stage above (at [8(c)]).

The second sale and purchase contract dated 30 May 2018 applied to the remaining three shipments of tin concentrate, ie, the seventh to ninth shipments. This contract states that the plaintiff was to take delivery as follows:

Again, this was the point of the “commencement of the transit” within cl 8.1 of the Policy (see [6] above). From those terms, it is undisputed that the insurance coverage for the tin concentrate of the seventh to ninth shipments began (on a free carrier basis) upon clearance of customs at the Bonded Warehouse in Kigali, Rwanda. This meant that the Transit Period for the seventh to ninth shipments began from the fifth stage above (at [8(e)]). The details of these shipments are as follows:

(a) For the seventh shipment, there were 30 drums of tin concentrate weighing 20.4 metric tonnes.

(b) For the eighth shipment, there were 36 drums of tin concentrate weighing 22.675 metric tonnes.

(c) For the ninth shipment, there were 40 drums of tin concentrate weighing 25 metric tonnes.

In total, under the sixth to the ninth shipments, the plaintiff was to take delivery of 86.075 metric tonnes of tin concentrate in 133 drums. For these shipments, the defendant issued marine cargo insurance certificates.

On or around 10 July 2018, the sixth shipment arrived at MSC in Penang, Malaysia. The seventh shipment was delivered to MSC on 25 July 2018. The eighth and ninth shipments were also delivered to the same place in Penang on 6 September 2018. Upon the arrival of the respective shipments, it was discovered that the entire quantity of the tin concentrate had been replaced with iron oxide. Upon inspection by MSC and representatives of both the plaintiff and the defendant, it was revealed that the top portion of the drums had been cut and welded. The welding marks around the rim of each drum was well concealed with a thick layer of white alkyd paint. In this way, the welded hole of the drum with the security seal of Precintia clips remained intact.

Upon the discovery of the thefts of the tin concentrate, the plaintiff sent notices of claim for the shipments to the defendant by e-mail on 17 July 2018 (sixth shipment), 31 July 2018 (seventh shipment) and 24 September 2018 (eighth and ninth shipments).

On or around 3 October 2018, the plaintiff’s representative, Mr Tambawala, and the defendant’s expert, Dr Luigi Petrone (“Dr Petrone”) from AqualisBraemar LOC, travelled to Rwanda. Dr Petrone went to Excellent Mining’s premises and the Bonded Warehouse to conduct investigations into the transit of the sixth to the ninth shipments from Kigali to Dar es Salaam.

On 7 May 2019, the defendant informed the plaintiff that its claims for the sixth to the ninth shipments regarding the swapped tin concentrate were rejected as the defendant’s investigations revealed that the tin concentrate had been swapped with iron oxide at Excellent Mining’s premises. Consequently, the defendant alleged that the plaintiff did not have an insurable interest as the loss did not occur during the Transit Period. Thus, the sixth to the ninth shipments were not covered by the insurance policy. On 19 June 2019, the plaintiff made a formal demand for payment of its claims under the Policy by way of a letter, but the defendant has not made such payment to date.

On 30 August 2019, the National Public Prosecution Authority of Rwanda (the “NPPA”) issued a report after conducting extensive investigations (the “NPPA Report”). It concluded that the thefts did not occur in Rwanda. The parties agree that such a report was issued. The defendant’s expert, Dr Petrone, however, did not accept the findings of the NPPA Report.

The parties’ cases

The plaintiff’s case

The plaintiff claims that the thefts of the tin concentrate in the sixth to the ninth shipments occurred during the Transit Period from Kigali, Rwanda to Penang, Malaysia, via Dar es Salaam, Tanzania.

The plaintiff relies on two important pieces of evidence to establish that the thefts took place when the drums of tin concentrate left the compound of Excellent Mining. These are:

(a) The fact that up till the point of customs clearance at the Bonded Warehouse in Kigali, none of the representatives stated above who had checked the tin concentrate and the sealed drums noticed any tampering with the sealed drums.

(b) The NPPA Report established that the thefts did not happen in Rwanda. The investigations included Excellent Mining’s premises, which was in Rwanda.

Accordingly, the plaintiff took out this Suit to claim for the full insured value of the sixth to the ninth shipments less a deduction of 0.5%, which amounts after deduction to US$1,154,508.94. In the alternative, the plaintiff claims for damages to be assessed.

The defendant’s case

The defendant denies that the thefts occurred during the Transit Period, ie, the thefts did not take place during the journey that was covered by the insurance policy. The defendant claims that the thefts of the tin concentrate occurred at Excellent Mining’s premises instead, ie, before the drums of tin concentrate began their long journey which was covered by the insurance policy. The defendant relies on the following evidence:

(a) Forensic evidence indicates the concomitant presence of cassiterite (tin oxide), coltan (columbite-tantalite) and wolframite (iron manganese tungstate) minerals (the “3Ts minerals”) in the iron oxide that was swapped for the tin concentrate. The 3Ts minerals are termed as such because they refer to tin, tantalum and tungsten. The presence of 3Ts minerals is significant because of the following reasons:

(i) Based on forensic analysis and the defendant's investigations, it is very likely these 3Ts minerals originated from mines in Rwanda as Rwanda is an exporter of these minerals. Tanzania produces only a very small amount of cassiterite (ie, a mineral containing tin) and does not produce coltan (ie, a mineral containing tantalum) and wolframite (ie, a mineral containing tungsten).

(ii) Iron oxide is a typical product of soil weathering and is available in tropical regions such as Rwanda, Tanzania and many parts of Africa. Alternatively, the iron oxide which was swapped for the tin concentrate may have been a by-product of the mineral beneficiation process of the 3Ts minerals. Hence, the defendant claims that there would have been no major difficulty in obtaining the requisite amount of iron oxide for the thefts in Rwanda.

(iii) It is likely that grains of the 3Ts minerals would be present in Excellent Mining’s premises because Excellent Mining as an exporter would purchase 3Ts minerals from various mines in Rwanda. The grains of all 3Ts minerals would have been admixed with the pile of iron oxide at Excellent Mining’s premises in the process of the pilferers filling the drums with the iron oxide to effect the swap of the tin concentrate.

(b) Forensic evidence also indicates that there was a very strong correlation between the white alkyd paint applied on the top and the rim of the drum lids. The paint applied by Excellent Mining’s personnel on the top of the drums was manufactured and commercially available only in Rwanda. It is likely that the same paint was applied by the persons who opened the drums and swapped their contents. The lids were welded back onto the drums, and the paint was applied thickly on the rims to hide the welding marks.

(c) There were windows of opportunity to swap the tin concentrate with iron oxide at the Excellent Mining premises. The sealed drums containing the tin concentrate were stored at Excellent Mining’s premises for a period of four to eight days, depending on the shipment, before the sealed drums were loaded into the 40ft container for transportation by Bolloré Logistics from Excellent Mining’s premises to Bolloré Logistics’ Bonded Warehouse. There were welding equipment and forklifts at Excellent Mining’s premises that could have been used to carry out the thefts. Dr Petrone, the defendant’s expert, alleged that the thefts could only have taken place at Excellent Mining’s premises as the latter was involved.

The defendant also claims that it was unlikely that the thefts occurred during the journey from Kigali, Rwanda to the port in Dar es Salaam, Tanzania and during the journey from Dar es Salaam, Tanzania to Penang, Malaysia. The defendant relies on the following evidence:

(a) The distance between Kigali and Dar es Salaam is about 1500km by land. The journey takes about three to five days depending on traffic conditions, refuelling stops and stops at government designated stopover points. Container trucks on long haulage from Rwanda to Dar es Salaam are not permitted to travel during the night for security reasons.

(b) Checks by customs at the Rwanda-Tanzania border are stringent. Any discrepancy between the actual container seal numbers and the seal numbers stated in the cargo documents would result in the container trucks (and their cargo) being seized by the relevant authorities.

(c) Container seals documented by Alex Stewart, Bolloré Logistics and the RRA in respect of the respective 40ft containers (containing the drums of purported tin concentrate and used for the land carriage) were found to be intact when the drums of tin concentrate were offloaded from the 40ft containers into the 20ft containers at Dar es Salaam.

(d) The doors of the 20ft container for each shipment were sealed by Tanzanian Customs, Bureau Veritas and the representative of the ocean carrier at Dar es Salaam. At MSC’s premises, the 20ft containers were inspected to be in good condition and the seal numbers tallied with those indicated in the respective bills of lading. Precintia wire clip seals spot-welded on the bung holes of the drum lids were observed to be intact at MSC’s premises.

The defendant also disputes the findings of the NPPA Report, citing the following reasons:

(a) No evidence was tendered on the composition of the iron oxide (containing 3Ts minerals) which was swapped for the tin concentrate.

(b) No evidence was tendered to show that the lids of the drums containing the tin concentrate were welded, and that thick paint was used to hide this. Further, no evidence was tendered to show that the paint used is only manufactured and commercially available in Rwanda.

(c) No evidence was tendered to show that the drums containing the tin concentrate were left at Excellent Mining’s premises for a period of four to eight days prior to being loaded into the respective containers for transportation by Bolloré Logistics from Excellent Mining’s premises. The fact that the drums were left at Excellent Mining’s premises for a period of time would have presented an opportunity for the pilferers to swap the tin concentrate. Further, welding equipment and forklifts were available at Excellent Mining’s premises to facilitate the thefts.

(d) No evidence was tendered to set out the circumstances of the land carriage of the tin concentrate from Kigali to Dar es Salaam, or of the offloading of the drums of tin concentrate at Dar es Salaam before they were loaded onto the respective vessels for delivery to MSC in Penang.

(e) The findings of the NPPA Report are inconclusive.

While the defendant denies its liability to pay the plaintiff, it does not dispute the quantum of full insured value in the plaintiff’s claim.

The sole issue to be determined from the agreed facts

At the conclusion of the trial, the following important facts were undisputed:

(a) In respect of the sixth to the ninth shipments, tin concentrate contained in the drums were stolen and were swapped with iron oxide.

(b) The thefts took place after the tin concentrate was loaded into the drums and the holes on the drums’ lids were welded and sealed with Precintia clips. Hence, the thefts took place at some point starting from the time the tin concentrate was stored at Excellent Mining’s premises in Rwanda to the time it arrived at MSC in Penang.

(c) There is no direct evidence to show who committed the thefts and where the thefts were committed. No relevant person suspected that anything was amiss until the tin concentrate arrived in Penang.

(d) If the thefts took place during the Transit Period, the defendant would be liable to indemnify the plaintiff’s loss.

These agreed facts are crucial, and I shall refer to them in the course of my analysis below.

Hence, in the present dispute, there is only one issue to be determined: where did the thefts of the tin concentrate in the sixth to the ninth shipments occur?

I shall turn now to my analysis and findings.

My decision

The applicable law

The parties agree that the burden of proof lies with the plaintiff to show that its loss, ie, the thefts of the tin concentrate, occurred during the Transit Period. This is simply a matter of trite law as he who asserts must prove.

However, the defendant takes a nuanced approach to the above position. The defendant submits that it is not required to prove on a balance of probabilities that the thefts occurred at Excellent Mining’s premises. Even if the defendant fails to convince the court that the thefts had occurred there, it does not immediately follow that the thefts occurred during the Transit Period. In other words, the court is not compelled to choose between the plaintiff’s and the defendant’s competing accounts. The court can hold instead that the plaintiff has failed to prove on a balance of probabilities that the thefts occurred after the commencement of the Transit Period.

For its position above, the defendant relies on the authority of Rhesa Shipping Company SA v Edmunds (The Popi M) [1985] 1 WLR 948 (“The Popi M”), affirmed locally in Clarke Beryl Claire (personal representative of the estate of Eugene Francis Clarke, deceased) and others v SilkAir (Singapore) Pte Ltd [2002] 1 SLR(R) 1136 (“Clarke Beryl Claire”) at [63], Surender Singh s/o Jagdish Singh and another (administrators of the estate of Narindar Kaur d/o Sarwan Singh, deceased) v Li Man Kay and others [2010] 1 SLR 428 (“Surender Singh”) at [121] and Wartsila Singapore Pte Ltd v Lau Yew Choong and another suit [2017] 5 SLR 268 (“Wartsila”) at [87].

In The Popi M, a ship sank in calm waters in the Mediterranean Sea. The owners of the ship sued the hull underwriters for the total loss of the ship arising from the sinking. For this loss to be covered by the insurance policy, the shipowners had to show that the ship sank as a result of the “perils of the sea” which would be covered by the policy. At first instance, the trial judge found that there was a collision with the ship which resulted in water entering through an aperture in the ship’s shell plating. The shipowners’ case was that this was because an underwater submerged submarine had collided with the ship. The hull insurer alleged that the aperture of the ship was due to the prolonged wear and tear of the ship’s hull over many years, ie, there was no collision involved. Having examined the relevant expert evidence, the trial judge rejected the hull insurer’s theory. In the circumstances, although he regarded the shipowners’ theory as “improbable”, he held that he was left with only the shipowners’ theory and accordingly allowed the shipowners’ claim. The hull insurer appealed and the Court of Appeal dismissed the appeal. The hull insurer further appealed to the House of Lords and it was held that the trial judge should not have regarded himself as being compelled to choose between the two theories by the parties. Rather, “[h]e should have borne in mind, and considered carefully in his judgment, the third alternative which was open to him, namely, that the evidence left him in doubt as to the cause of the aperture in the ship’s hull, and that, in these circumstances, the shipowners had failed to discharge the burden of proof which was on them” (at 956E). Hence, the House of Lords allowed the hull insurer’s appeal.

As stated in Surender Singh at [121], the principle in The Popi M was adopted by the Singapore Court of Appeal in Clarke Beryl Claire. Succinctly restated, the principle is that a claimant bears the burden of proof and if he fails to persuade the court that his case has been proven on the balance of probabilities, then judgment should be given for the defendant. The test is not whether the claimant’s case is more probable than the defendant’s, but whether the claimant’s evidence (and not hypothesis) has been proven on a balance of probabilities.

Generally, the above position undoubtedly applies to the present case. In this regard, I shall analyse the evidence that the parties relied on to ascertain whether the plaintiff has discharged its burden of proof. I should also add that I agree with the House of Lords that the trial judge had erred in his reasoning. Having regarded the shipowners’ reasoning as “improbable”, the trial judge should not have felt compelled to accept this theory.

However, there are important salient differences between The Popi M and the present case. In The Popi M, the cause of loss, viz, the cause of the aperture in the ship’s plating, was disputed. In the present case, the cause of loss is not disputed: the thieves cut open the drums’ lids and swapped the tin concentrate with iron oxide. What is disputed is the time (and location) that the thefts occurred. Conversely, in The Popi M, it was undisputed that the ship sank at sea.

Similar differences are also present between the rest of the cases the defendant relied on and the present case. In all the cases cited by the defendant above (at [30]), the cause of the loss suffered by the plaintiff was unknown. In Clarke Beryl Claire, the Court of Appeal held that “[t]he cause of the [aircraft’s] crash was not conclusively revealed by the technical evidence.”

In Wartsila, the cause of a vessel’s engine breakdown was disputed. While the plaintiffs maintained that the engine breakdown was due to Wartsila’s poor workmanship and/or negligence in carrying out the repairs to the vessel’s main engine, Wartsila averred that the engine breakdown was caused by the vessel’s operation and the crew’s maintenance of the main engine and/or supply and use of defective equipment and parts in the operation of the main engine.

In Surender Singh, the cause of the deceased’s death was disputed. The plaintiffs were required to prove on a balance of probabilities that the third defendant’s breach of duty in exercising post-operative monitoring of the deceased had caused or materially contributed to the deceased’s death from or substantially from blood loss.

Therefore, the facts of the above cases differ materially from the present case. In this case, the cause of the plaintiff’s loss is known and undisputed, ie, the plaintiff suffered loss due to theft.

Since the thefts of the tin concentrate are not disputed, the issue is simply where did the thefts occur. The resolution of this issue involves an almost binary choice for my consideration: either the thefts took place during the Transit Period or did not. For the sixth shipment the theft either took place at the Excellent Mining premises or elsewhere during the Transit Period. If the theft could not have occurred at the Excellent Mining premises, then the theft must have occurred during the Transit Period. For the seventh to the ninth shipments, the thefts either took place at the Excellent Mining premises or along the 45 minutes to one hour journey from Excellent Mining’s premises to the Bonded Warehouse (see [10] above), or it took place elsewhere during the Transit Period. However, once it is both shown that the thefts were unlikely to have occurred at Excellent Mining’s premises and this short journey thereafter, then it follows that the thefts had occurred during the Transit Period. After all, the parties agreed that the thefts did occur and that the thefts must have occurred at some point during the whole journey from Excellent Mining’s premises in Rwanda to MSC in Penang (see [24(b)] above).

My view above is supported by Lord Brandon’s reasons for rejecting the trial judge’s reasoning in The Popi M. He stated as follows (at 955G to 956D):

In respect of the first reason, I accept that I am generally not always bound to make a finding one way or the other with regard to the facts averred by the parties.

In respect of the second reason, Lord Brandon disapproved of the method of eliminating possibilities to determine the true state of affairs (ie, a “process of elimination”), coined by the fictional character, Mr Sherlock Holmes, and adopted by the trial judge. His Lordship stated that a process of elimination could not be used where there are possible unexplored explanations for the cause of loss. Indeed, as stated above, “the ship sank in such deep water that a diver’s examination of the nature of the aperture, which might well have thrown light on its cause, could not be carried out”. Hence, in The Popi M, the cause of loss was indeterminate. In fact, although the trial judge in The Popi M accepted the shipowners’ submarine theory, he regarded it as extremely improbable. In other words, a submarine collision and prolonged wear and tear were not the only possible causes that could explain the presence of the aperture which sank the ship. The ship could have, for instance, been deliberately sabotaged for some nefarious reason. This is why Lord Brandon placed much weight on the fact that a diver’s examination could not be carried out: such an examination could have suggested other causes for the aperture leading to the ship’s sinking.

In the present case, however, the cause of loss and the modus operandi of the thefts were agreed, ie, the thefts occurred by swapping the tin concentrate with iron oxide. Although the time and location of the thefts were disputed, the fact remains that the thefts had occurred at some point in the whole journey. As stated above (at [24(b)]), the parties agreed that the thefts must have occurred at some point from the time the tin concentrate was loaded into the drums and sealed at Excellent Mining’s premises in Rwanda to the time it arrived at MSC in Penang. Accordingly, the various stages in this journey agreed by the parties represent a self-contained set of possibilities. In other words, the thefts must have occurred at some point within this circumscribed period of time. Where one option is shown to be unlikely, the likelihood of the other option goes up significantly. Hence, it is precisely the case here that all relevant facts are known such that all possible explanations, except an extremely improbable one, can properly be eliminated. In this way, on the authority of The Popi M itself, the process of elimination is permitted in the present situation.

In respect of the third reason, I agree that the legal concept of proof of a case on a balance of probabilities must be applied with common sense. However, ex facie, the present case does not involve evidence which would “leave [me] in doubt whether the event occurred or not”. The parties agreed that the thefts had occurred at some point from the time the tin concentrate was loaded into the drums and sealed at Excellent Mining’s premises in Rwanda to the time it arrived at MSC in Penang. It is undisputed that the thefts had occurred. The central issue is where did the thefts take place.

Hence, Lord Brandon’s reasons for rejecting the trial judge’s method of using a process of elimination to determine the true state of affairs do not apply in the present case as the facts here are materially different.

However, I wish to explain how the process of elimination would operate in the present case. This is important because the defendant submits that: (a) a process of elimination entails reversing the standard of proof from a balance of probabilities to a balance of improbabilities (see Hua Seng Sawmill Co Bhd v QBE Insurance (Malaysia) Bhd [2003] 4 SLR(R) 449 at [68]); and (b) a process of elimination would arrogate the burden to the defendant to prove the truth of an alternative cause (see Hub Warrior Sdn Bhd v QBE Insurance (Malaysia) Bhd [2004] SGHC 279 at [52]).

In The Popi M, the House of Lords noted (at 954F) that “[the trial judge] had before him a mass of expert evidence relating to the possibilities that the proximate cause of the ship's loss was a collision with a submerged submarine on the one hand or wear and tear of the shell plating on the other.” However, although such expert evidence was adduced, the trial judge was not convinced by either party’s case theory.

Here, the plaintiff has adduced evidence (direct, circumstantial and forensic evidence) in support of its submission that: (a) it was unlikely that the thefts occurred at Excellent Mining’s premises (and the immediate journey after); and (b) it was probable that the thefts had occurred during the Transit Period. Given that it is undisputed that the thefts had to have taken place at some point in the whole journey of the transportation of the tin concentrate, I would consider that the plaintiff’s evidence on both points, ie, (a) and (b) above, goes towards discharging its burden to prove its case on a balance of probabilities. The plaintiff referred the court to credible evidence such as the NPPA Report and the evidence of tight and layered security at Excellent Mining’s premises to show that it was highly unlikely for the thefts to have occurred at the Excellent Mining premises. This evidence serves to diminish the defendant’s case theory that the thefts took place in the premises of Excellent Mining and at the same time the evidence bolsters the plaintiff’s case that the thefts must have occurred at some point during the Transit Period. This case is unlike the facts of The Popi M where there could have been an unknown cause for the aperture in the ship’s hull beyond those postulated by the parties that the parties may not have considered, such as sabotage (see [42] above).

Hence, the present inquiry is significantly different from that in The Popi M. The court is not compelled to choose one party’s theory over another by a simple comparison of their relative likelihood, however improbable that theory may be. It is not a case where the burden has shifted to the defendant to prove the truth of an alternative cause. Rather, I have to consider whether the plaintiff’s submissions and supporting evidence have proven its case on a balance of probabilities. There is, therefore, no departure from the established trite principles governing the plaintiff’s burden and standard of proof in the present case.

Also, I should add that in this case, there is no direct evidence of the thefts. Thus, I have to rely on indirect, circumstantial and forensic evidence to ascertain whether the plaintiff has proven its case on a balance of probabilities.

I shall now refer to my analysis.

The level of security during various points in the transport of the tin concentrate

I have set out the various stages in the transport of the tin concentrate above (at [8]).

Excellent Mining’s premises

I shall first turn to the security at Excellent Mining’s premises. To recapitulate, the defendant claims that the thefts had occurred there.

The plaintiff submits essentially that since Excellent Mining’s premises were under tight security, it was unlikely that the thefts had occurred there.

Mr Theodore Sindikubwadbo (“Mr Sindikubwadbo”) was Excellent Mining’s director at the material time. He testified that the level of security at Excellent Mining’s premises was very high.

Mr Sindikubwadbo stated that the premises were secured by compound walls and fences. The compound walls were fitted with thorn fences to prevent unauthorised entry. There was also a metal gate at the entrance which would be locked at the end of every day. There were also CCTV cameras at Excellent Mining. Each recording would remain for ten days before it was overwritten by new footage.

At the time when the drums of tin concentrate were stored on Excellent Mining’s premises, four security guards were deployed. The guards had to, inter alia, patrol the premises around the clock in shifts to ensure that: (a) there was no unauthorised entry; and (b) the drums of tin concentrate were securely stored on the premises. If there were any signs of tampering with, or theft of, the drums of tin concentrate, the guards were required to immediately report such an incident to Excellent Mining’s management. There were no such reports during the relevant period. Moreover, at the material time, Mr Sindikubwadbo and his family were staying at a house adjacent to the premises. In that way, he could easily ensure that the guards were working. Further, according to Mr Sindikubwadbo, because Excellent Mining’s premises are not large (measuring approximately 569m2), it would be very unlikely that unauthorised entry would go unnoticed, so four guards were sufficient to guard it.

Mr Sindikubwadbo further stated that after the bags of tin concentrate which were sealed with Precintia clips arrived at Excellent Mining’s premises, they were emptied onto the ground. The MFO stationed at Excellent Mining, who was a representative of the RMB (see [8(a)] above), would check to ensure that the Precintia clips on the sealed bags match the contents of those bags. These clips stated, inter alia, the weight, source and origin of the tin concentrate. The MFO collected all the clips from the bags of tin concentrate and placed them into the metal box which was welded to the top of one of the drums that contained the tin concentrate. It was after the bags of tin concentrate were emptied and mixed that a representative of Alex Stewart would collect a sample of the tin concentrate for analysis. Thereafter, it would take about 16 men and seven hours, ie, one full day, to fill the drums with the tin concentrate.

The MFO, who was stationed at Excellent Mining, along with representatives from ITSCI, conducted routine spot checks throughout the day on the drums of tin concentrate. The MFO would inspect the drums of tin concentrate before they left Excellent Mining’s premises for the day. When they returned the following morning, they checked the sealed drums of tin concentrate again to ensure that no thefts had occurred during the night.

The evidence of Mr Felicien Nkomeje (“Mr Nkomeje”), a Certification Specialist at the RMB, corroborates the above evidence. According to Mr Nkomeje, at the material time, the MFO was Ms Campire Laurence (“Ms Laurence”), and she was stationed at Excellent Mining’s premises permanently. In other words, Ms Laurence was at Excellent Mining’s premises from 7.00am to 5.00pm every day. Mr Nkomeje also testified that Excellent Mining’s activities relating to minerals could not continue without Ms Laurence’s supervision. Part of Ms Laurence’s duties was to check for signs of tampering of the sealed drums containing the tin concentrate, and she had to do so daily. Therefore, if the lids of the drums were cut and the contents of the drums were replaced, Ms Laurence would have noticed such tampering. Ms Laurence would then be required to raise this concern immediately to the RMB’s office and to ITSCI. However, no such issues were raised during the material time.

According to Mr Sindikubwadbo, during the period when the sealed drums of tin concentrate were stored at Excellent Mining’s premises, there were no signs of tampering or theft observed.

The evidence of Mr Norman Mwashi (“Mr Mwashi”), the Managing Director and Chief Chemist of Alex Stewart, largely corroborates Mr Sindikubwadbo’s evidence of the above. Mr Mwashi also testified that all the requisite inspections were done before and after the sealed drums of tin concentrate were transported from Excellent Mining’s premises to the Bonded Warehouse. No signs of tampering were observed. Indeed, the defendant’s expert, Dr Petrone, opined that the people involved in checking the seal of the drums were not involved in the thefts.

From the above evidence, it is clear that the drums of tin concentrate were under tight security while it was in Excellent Mining’s premises.

I do note, however, that Mr Mwashi gave evidence that the loading of the drums into the 40ft container occurred outside Excellent Mining’s premises. This was because the truck was too large to enter the premises. The loaded container would remain outside till the clearance papers were ready. Nevertheless, the container would still be under surveillance.

Dr Petrone, however, opined that Excellent Mining’s premises was “the best place” to carry out the thefts because it had the facilities to carry out the swapping, such as cutting/welding equipment, excavators, manpower and white paint. He even suggested that for such a massive operation to take place, Excellent Mining must have been complicit in the thefts. Yet, these were merely Dr Petrone’s bare assertions without any supporting evidence or proof.

In contrast, the plaintiff’s expert, Mr Howard Nathan Wheeler (“Mr Wheeler”), came to a completely different conclusion. He mentioned in his report the extensive difficulties that the thief would have encountered to swap the tin concentrate with iron oxide at Excellent Mining’s premises:

In addition to the above considerations, I also note that there are CCTV cameras in Excellent Mining’s premises. Mr Nkomeje testified that he saw such surveillance cameras there. The defendant also produced a photograph showing one such camera at the premises. However, I note that Mr Nkomeje did not know which particular areas were being monitored by the CCTV surveillance cameras.

Having carefully considered the evidence above, even if the necessary equipment were available, it was clearly logistically impossible for the thefts to have taken place at Excellent Mining’s premises, especially with the several layers of checks and surveillance at Excellent Mining. Dr Petrone knew this. At the trial, he acknowledged that his theory that the thefts took place at Excellent Mining was fundamentally premised on Excellent Mining’s staff being complicit in the thefts. Otherwise, his theory would collapse. I set out his testimony in court below:

The sheer amount of manpower and time needed, the truck load of iron oxide required, the additional truck needed to take away the tin concentrate, the substantial amount of noise that would have been generated, and the high level of security in the premises make such a plan unfeasible without the involvement of Excellent Mining.

From Excellent Mining’s premises to the Bonded Warehouse

As previously noted (see [8(c)] above), the journey from Excellent Mining’s premises to the Bonded Warehouse takes about 45 minutes to one hour. Even if there was inadequate security along this journey, there was plainly insufficient time and space for thefts of the present magnitude to occur. The parties did not suggest that the thefts took place during this short stretch of the journey.

Relevant personnel had also inspected the sealed drums at the Bonded Warehouse (see [8(d)] above). According to Mr Nkomeje, the inspection at the Bonded Warehouse would have had to be carried out on the same day as the delivery.

From the Bonded Warehouse to the port in Dar es Salaam

Dr Petrone discounted any possibility of the thefts having occurred during the journey from the Bonded Warehouse to Dar es Salaam. His opinion in relation to security focused on the fact that the seals on the container doors were intact:

However, Mr Wheeler disagreed with Dr Petrone’s reasoning that no one gained access into the sealed container because the seals of the container’s doors were intact. Mr Wheeler stated as follows:

At the trial, Mr Wheeler was questioned on this portion of his report, and he testified as follows:

In my view, Mr Wheeler’s opinion is reasonable and credible. The parties agree that the thefts were committed by circumventing the seals on the top of the drums’ lids, leaving the sealed holes of the drums untouched. There was hardly any sign of tampering at the cut and welded portion of the drums’ top as well. The thefts were therefore professionally and skilfully done. When the sealed drums arrived at MSC in Penang, the defendant’s experts, namely Mr Desmond Sim Kok Whye (“Mr Sim”), the General Manager of Zama Marine Services and Consultancy Sdn Bhd (“Zama”), and Mr Yeoh Oon Huat (“Mr Yeoh”), an ad-hoc surveyor with Zama, inspected the drums. Mr Yeoh and his assistant, Mr Mohan a/l Govindasamy, who were Mr Sim’s surveyors on site, could not even spot the tampering. Mr Sim and Mr Yeoh concluded that the sealed drums were untampered. In this regard, Mr Sim testified as follows:

Mr Yeoh testified as follows:

Clearly, the swaps were not done by amateur thieves but by a professional and well-organised gang of thieves. If such professional thieves could cut and weld the top of the drums to circumvent the security seals to gain access to the tin concentrate without leaving any visible trace of tampering, they would have no problem in using the same modus operandi to gain access to the sealed drums in the sealed container without tampering with the seals of the container’s doors. Dr Petrone’s expert report does not indicate that he had examined the state of the 40ft containers.

More importantly, as noted above (at [8(e)]), the parties agreed that the 40ft container with the drums of tin concentrate travelled overland from the Bonded Warehouse in Kigali, Rwanda to the port in Dar es Salaam, Tanzania. This journey was about 1400km to 1500km and it required several overnight stops.

The defendant appointed Mr Sim to investigate the loss of the tin concentrate in the present matter. Mr Sim claimed that he had a telephone conversation with one Mr Alphonce Emmanuel (“Mr Emmanuel”), the Operations Officer of Bolloré Transport and Logistics Tanzania Ltd. Mr Sim claimed that Mr Emmanuel told him the following details about the supply chain:

(a) Travel was only during the day with no intermittent halt at any point except for refuelling. Container trucks on long haulage from Rwanda to Dar es Salaam are not permitted to travel during the night for security reasons.

(b) Overnight halts were only allowed at government-designated stopover points. These stopover points were manned by security personnel who carried out checks on all container trucks that passed.

(c) Should there be any discrepancy found between the seal number on the container’s doors and the travel documents, or if any tampering of the container or seals is noted, the container truck and its cargo would be seized.

In my view, even assuming that the above details are true, they do not imply that the security measures were watertight as the security measures were only at the designated spots. The thefts could have occurred along any place during the long stretch of the Transit Period which would be away from anybody’s notice. The nefarious operation only had to compromise the container’s driver and his assistant, if any. Indeed, Mr Sim acknowledged that “no security was provided to the conveyance vehicle or the drivers”. Moreover, as Mr Wheeler states in his report, “the 6 to 17 days transit time [from] Kigali to Dar [e]s Salaam appears to offer less supervision [ie, as compared to that in Excellent Mining’s premises] and as much, if not more, time and within its 1,400 km it seems reasonable to presume that there would be many areas that would allow a truck to stop and cargo operations to be undertaken without raising suspicion.”

As regards the port in Dar es Salaam itself, Mr Wheeler opined that it was possible for the thefts to occur there too. He testified that port areas are generally chaotic places, which suggests that thefts would be difficult to spot. Also, according to him, graft was commonplace in Tanzania, such that 10% of goods there are “under peril of theft”. I find that such evidence is speculative and I place little weight on it. Compared to the journey to the port, the port itself would naturally have the presence of more people in the vicinity. Indeed, according to the defendant, there were CCTV cameras and security guards at the port in Dar es Salaam.

From Dar es Salaam to Penang

The parties’ experts, especially the plaintiff’s, did not opine that it was likely for the thefts to have occurred in the remaining stages of the Transit Period. Hence, it is unnecessary for me to analyse them. I note briefly that it would have been very difficult for the thefts to have occurred on the voyage to Penang and at the port in Penang itself. As a matter of common sense, there would have been too many people around for thefts of such magnitude and sophistication to take place unnoticed.

Finally, to dispel the notion that the determination of where the thefts took place involves a binary consideration, as I have alluded to above at [39], the defendant raised in oral submissions that it was possible for the thefts to have occurred at MSC in Penang. I note that this suggestion is contrary to Dr Petrone’s evidence, where he agreed it was not his case that the thefts could have occurred at MSC in Penang (see [137] below).

I, therefore, find that the defendant’s belated suggestion that the thefts could have occurred at MSC in Penang was of no significance.

Conclusion on the level of security

Examining the above evidence in totality, from a comparison of the level of security at different stages, the weakest link in the entire journey of transporting the tin concentrate must have been the journey from the Bonded Warehouse to the port in Dar es Salaam. The immediate logical observation is that the weakest link exposes numerous opportunities to the thieves to perpetuate the thefts without being detected as compared to the Excellent Mining premises and the locations in the rest of the Transit Period. Furthermore, the six to 17 days’ road journey offers more space and time for the necessary logistics to be carried out skilfully.

I also note that, at the trial, Mr Sim could not provide an adequate counterargument to the above analysis. For clarity and completeness, I reproduce his long explanation below:

Evidently, even Mr Sim himself thought that it sounded “bizarre” for him to claim that the weakest link in the whole journey was Excellent Mining’s premises, given the tight security there.

From Mr Sim’s opinion he appears to be partial and lacks objectivity in his support of the defendant’s case.

To begin with, Mr Sim was made aware of the insurance coverage by the defendant for the sixth to the ninth shipments as early as 2018. He knew that if the thefts took place at Excellent Mining’s premises the defendant would not be liable. The thefts of these shipments occurred in 2018. Why was Mr Sim informed of the insurance coverage? If he was commissioned to do an independent investigation into the thefts, there was no need for him to be apprised of such a fact which would likely have affected the impartial conduct of his investigation.

Moreover, Mr Sim claimed that he did not know the lids of the drums were cut. This opinion goes against what the parties had agreed on, ie, the lids of the drums were cut to swap the tin concentrate with iron oxide. This called into question the very purpose of his investigation, which was then posed to him:

I find Mr Sim’s explanation to be plainly illogical. He opined that there were no thefts and that the drums were not cut. These findings are against the very crux of the parties’ agreed facts, ie, it is undisputed that there were thefts of the tin concentrate and that the thieves cut the top of the drums to gain access to the tin concentrate and swapped it with iron oxide. It seems that Mr Sim was trying to tailor his evidence in the defendant’s favour, to the extent that it would – in his own view – sound “bizarre” to a reasonable person (see [85] above).

Hence, from the above considerations and in the absence of credible evidence or explanation from the defendant, I find that the weakest link in the whole journey must have been the journey from the Bonded Warehouse to the port in Dar es Salaam for the thefts to have occurred for the sixth to the ninth shipments.

The NPPA Report

The NPPA Report dated 30 August 2019 was released after investigations by the Rwanda Investigation Bureau (“RIB”). The NPPA was satisfied that the thefts had not been committed in Rwanda and decided to close the case. In the NPPA Report, Excellent Mining, represented by Mr Sindikubwadbo, and Bolloré Logistics, represented by Mr Bisangwa François, were identified as suspects for the charges of breach of trust and theft. As noted above (at [16]), the parties agree that such a report was issued.

In brief, the findings of the NPPA Report are as follows:

(a) The investigations showed that the tin concentrate was genuine, and it was loaded into drums and sealed in the presence of staff from the RRA, Alex Stewart, and the RMB.

(b) The plaintiff produced a report from Bureau Veritas, which conducted the inspection on the containers loaded with the drums of tin concentrate. The report clarified that no sabotage was done to the containers and the drums from Kigali to Dar es Salaam, as the drums, containers and tags were in “good condition”. Hence, it “prove[s] that no crime of breach of trust or theft has been committed in Rwanda.”

(c) The plaintiff produced a report made by Alex Stewart, which was obligated to conduct “professional inspection and analysis of minerals, to confirm their nature, quantity and quality and to confirm the nature of exported minerals as well as their quantity.” At the time of loading the tin concentrate into the drums, a representative of Alex Stewart was present and confirmed the nature, quality and quantity of the minerals to be in conformity with those agreed by Excellent Mining and the plaintiff. Consequently, on the basis of this report, “there is no doubt that no crime of breach of trust or theft was committed in Rwanda”.

(d) All the parties that were interrogated, viz, Excellent Mining, the plaintiff and Bolloré Logistics, did not “confirm” that the minerals were stolen in Rwanda.

I pause to emphasise that the findings of the Report are expressly limited to Rwanda only (see [91(c)] above), ie, whether Excellent Mining and Bolloré Logistics were complicit in the thefts in Rwanda. The Report does not cover other stages in the transport of the tin concentrate thereafter, ie, the journey from the Bonded Warehouse in Kigali, Rwanda to the port in Dar es Salaam, Tanzania. In this regard, the defendant submitted erroneously that the NPPA “made a finding” that “there was no ‘sabotage’ on the drums and containers from Kigali to Dar [e]s Salaam.” As stated above (at [91(b)]), that was a finding by Bureau Veritas based on which the NPPA made a finding that “no crime of breach of trust or theft has been committed in Rwanda” [emphasis added].

However, the defendant claims that the above findings are inconclusive for the following reasons:

(a) The checks performed by the various authorities and Alex Stewart to ensure that the tin concentrate loaded into the drums was genuine are effective only in so far as the drums were not subsequently opened after sealing but prior to delivery to MSC.

(b) Sampling and testing of the contents in the drums were not likely to have been carried out by Bureau Veritas during the loading into the 20ft container at Dar es Salaam as the Precintia clip seals spot-welded on the bung holes of the drum lids were observed to be intact at MSC’s premises. The defendant claims that Bureau Veritas would have limited its checks to “checking that the Precintia wire clip seals were intact and/or counting/weighing the drums and/or checking the Certificate of Assay and/or Sampling, Weighing and Packing certificates.”

(c) Although Excellent Mining, Bolloré Logistics and the plaintiff did not confirm that the tin concentrate was stolen in Rwanda, it does not necessarily follow that the swap of tin concentrate did not in fact happen in Rwanda.

The defendant and its expert, Dr Petrone, do not accept the NPPA’s findings and the above opinions were a rebuttal to the NPPA Report’s findings. However, the substance of the defendant’s dispute with the NPPA’s findings is that, in its view, the NPPA had simply conducted a “desktop review of reports by different surveyors” and it did not discover the method of thefts. In particular, the defendant argues that the NPPA Report relied on the report of Bureau Veritas and its “suspect” finding of fact that the drums were not tampered with, in order to conclude that there was no evidence to prove that a crime of breach of trust or theft had been committed in Rwanda. This is consistent with the defendant’s pleaded criticism in its Defence (see [22(a)]–[22(d)] above). In cross-examination, Dr Petrone explained as follows:

To begin with, the NPPA is Rwanda’s prosecution authority. The NPPA clearly mentioned in the Report that the investigations were conducted by the RIB and the NPPA upon the complaint lodged by the plaintiff’s representative, Mr Tambawala. The ambit of the NPPA’s investigations would, as a matter of logic, be limited to whether the thefts had occurred in Rwanda. If the NPPA found that there were no thefts in Rwanda, it would not have been necessary for it to further investigate the method of thefts, which would have happened elsewhere.

Moreover, Dr Petrone had no basis to allege that the NPPA had conducted an improper investigation as he did not know the extent of investigation into this case that was performed by the NPPA and the RIB. In cross-examination, Dr Petrone gave the following unsatisfactory answers:

As can be seen from the above, Dr Petrone essentially admitted that his basis for claiming that the NPPA had done no “independent” investigation on this matter was confined to the NPPA Report itself. Dr Petrone was not aware of what investigations were carried out apart from the brief summary in the NPPA Report. Ironically then, it was Dr Petrone who had no independent basis for his criticism. From Dr Petrone’s testimony, it is clear that he completely rejected the NPPA Report without any reasonable and persuasive basis. Dr Petrone’s purported investigation into this case when he was in Rwanda with Mr Tambawala was highly unsatisfactory, inadequate, biased and lacked objectivity. I shall elaborate on his investigation when he was in Rwanda below.

It is unfortunate that the plaintiff did not call the official from the NPPA in charge of the investigation into this case to testify in court. Therefore, some caution is warranted when relying on the NPPA Report. Although the defendant did not object to the admissibility of the NPPA Report, I am aware that the defendant had serious criticisms of the NPPA Report. However, the defendant has not produced reasonable or credible evidence for me to jettison the findings of the NPPA Report. It is, therefore, inappropriate and unreasonable for me to ignore the findings made by the NPPA on the basis of Dr Petrone’s unjustified opinion. I also reiterate that the findings of the NPPA are limited to whether Excellent Mining and Bolloré Logistics were complicit in the thefts in Rwanda (see [92] above). The defendant’s counsel in his oral submissions accepted that the jurisdiction of the NPPA is limited to Rwanda. The findings, therefore, do not preclude the possibility of the thefts occurring during the journey from beyond Rwanda to Dar es Salaam, Tanzania, and the other stages in the transport of the tin concentrate.

The 3Ts minerals and the swapped iron oxide

The defendant claims that there is forensic evidence which indicates the concomitant presence of cassiterite (tin oxide), coltan (columbite-tantalite) and wolframite (iron manganese tungstate) minerals (ie, the 3Ts minerals) in the iron oxide that was swapped for the tin concentrate (see [20(a)] above).

According to the defendant, through Dr Petrone, the presence of 3Ts minerals in the swapped iron oxide weighs in favour of the finding that the thefts occurred at Excellent Mining’s premises. It submits that the replacement iron oxide was kept in a place that processes the 3Ts minerals before it was loaded into the drums. As a result, the 3Ts minerals were eventually found together with the iron oxide in the drums. In this regard, the defendant claims the following:

(a) Excellent Mining processes the 3Ts minerals, so the above forensic evidence shows that it is likely that the iron oxide was stored at Excellent Mining’s premises. By extension, the thefts would have happened there.

(b) The iron oxide could not have been stored in Tanzania. While Tanzania has some mines with cassiterite (ie, a mineral containing tin, one of the 3Ts minerals), it does not have mines with wolframite (ie, a mineral containing tungsten, one of the 3Ts minerals) and coltan (ie, a mineral containing tantalum, one of the 3Ts minerals).

The defendant’s experts are Dr Hans-Eike Gäbler (“Dr Gäbler”), a chemist at the Federal Institute for Geosciences and Natural Resources in Germany, and Dr Petrone, who is a Principal Scientist at AqualiasBraemar LOC. When Dr Petrone went to Penang from 12 to 13 September 2018, he took four samples of the materials in the drums that arrived in Penang in the presence of Mr Tambawala, the Managing Director of the plaintiff, and a surveyor from Alex Stewart International (Malaysia). He then sent these samples to Dr Gäbler on 20 February 2019 and these samples were tested on 28 February 2019, 1 March 2019 and 4 March 2019.

Dr Gäbler was instructed by Dr Petrone to check: (a) whether the cassiterite found in the four samples came solely from two mine sites in Nyaruvumu and Gituntu; and (b) if all four samples came from the same source. Dr Gäbler prepared two reports, in which he concluded as follows:

(a) It was improbable that the cassiterite grains in one sample originated from the two mine sites in Nyaruvumu and Gituntu.

(b) Instead, the age estimation of the cassiterite grains in the aforementioned sample showed that the Kibara Belt (which is in east-central Africa) and some other locations of the same age may be the origin of the cassiterite while many other locations worldwide can be excluded as they are different in age.

(c) It is “very plausible” that the cassiterite grains in all four samples originate from the same source.

With regard to (a) and (b), these two provinces in Rwanda were near Kigali. Dr Petrone explained that he identified these two regions because he was informed by the defendant that the origin of the materials sampled was from these two mines.

With regard to (c), Dr Gäbler clarified that a source did not mean a specific geographical region or a specific mine. Rather, his method of analysis, the Analytical Fingerprint (AFP) method, relies on comparing the geochemical signature of the samples with that of a reference material. Different mines would have different signatures. However, where material from one mine is combined with material from another, that mixture has its own signature, which is in his view also a source.

Dr Gäbler also clarified that he was not instructed to ascertain the origin of the other materials in the samples given to him. He explained that this was because he did not have a database to compare the samples to. Dr Petrone confirmed that Dr Gäbler told him that it was not possible to ascertain the origin of the coltan (columbite-tantalite), wolframite (iron manganese tungstate) and iron oxide.

I find it difficult to accept that Dr Gäbler’s report assists the defendant’s case. It is undisputed that the origin of the cassiterite, ie, the tin concentrate, was from Excellent Mining’s premises. Dr Gäbler did not analyse where the other two 3Ts minerals and the iron oxide originated from, which would have been pertinent to the present issue.

On the basis of Dr Gäbler’s report, Dr Petrone makes the following findings in his report:

While Dr Gäbler’s report concerned only cassiterite, Dr Petrone’s findings concern the 3Ts minerals as a whole. Essentially, he claimed that since (a) the 3Ts minerals are rarely found together in the same mine, (b) Excellent Mining trades in the 3Ts minerals, and (c) the samples did contain the 3Ts minerals, it means that it was likely that the 3Ts minerals were from Excellent Mining’s premises. Hence, he inferred that the 3Ts minerals which were “present on the floor of the open yard or on other equipment/tools used at Excellent Mining” were admixed with the iron oxide prior to the drums being filled. Dr Petrone also referred to a photograph he took which showed a red substance in Excellent Mining’s premises. On this basis, he claimed that Excellent Mining had iron oxide at its premises. Hence, the iron oxide was mixed with the 3Ts minerals at Excellent Mining’s premises prior to loading into the drums. He therefore concluded that the thefts occurred at Excellent Mining’s premises.

I shall first address Dr Petrone’s claim that iron oxide was present at Excellent Mining’s premises. Dr Petrone did not take a sample of the red substance for testing. When questioned, he conceded that there could be other red substances that are not iron oxide. Hence, the photograph he referred to is inconclusive evidence. Yet, Dr Petrone was so quick to conclude assertively that the red colour in the photograph shown at the compound outside Excellent Mining is iron oxide. He agreed that to be certain that it was iron oxide an analysis of the red substance is necessary. But, nevertheless, he continued to insist that it was iron oxide.

I turn next to Dr Petrone’s theory that the presence of the 3Ts minerals in the drums meant that the swap took place at Excellent Mining’s premises. His theory is premised on the admixing of iron oxide with the 3Ts minerals prior to the swap of the tin concentrate with iron oxide.

Dr Petrone’s theory is not the only possible explanation for the presence of the 3Ts minerals. If the 3Ts minerals were found on the ground of Excellent Mining’s premises, it is possible that they were admixed with the tin concentrate when the tin concentrate was unloaded onto the ground before being loaded into the drums (see [8(a)] above). After the swap with iron oxide, there could be remnants of the tin concentrate and the 3Ts minerals left in the drums. When Dr Petrone was questioned at the trial, he conceded this point:

Similarly, the plaintiff’s expert, Dr Mirjana Küzma (“Dr Küzma”), a material scientist, testified that the presence of the 3Ts minerals could be explained by the remnants of the tin concentrate in the drums after the swap:

Furthermore, the drums used were not new: they were reused drums which might have contained the 3Ts minerals. Therefore, it does not matter where the swap took place as the 3Ts minerals could have been present in the drums in any case. Hence, Dr Petrone’s inference that the swap took place at Excellent Mining’s premises due to the presence of the 3Ts minerals is not conclusive but speculative.

However, Dr Petrone made an argument in support of his claim above that the 3Ts minerals must have been mixed with the iron oxide, instead of being admixed with remnants of the tin concentrate. He argued that the iron oxide filled the whole of the drums and thus there was little room for thorough mixing of the iron oxide in the drums. The remaining tin concentrate (after the swap) would be found mostly at the bottom of the drums. When Dr Petrone took the samples of the iron oxide in Penang, he did not take them from the drums’ bottom. Hence, he claimed that what he sampled was purely the iron oxide that was swapped, without the remaining tin concentrate that was originally in the drums. However, the iron oxide in the drums must have been subjected to considerable shaking and vibration during the long journey on land, when they were removed from the 40ft container to the 20ft container, and during the long sea voyage. Moreover, as Mr Wheeler testified, cassiterite is a “dusty and free-flowing” material. Thus, Dr Petrone’s explanation is not convincing. I therefore do not agree with Dr Petrone’s theory that the swap must have taken place at Excellent Mining’s premises because of the presence of the 3Ts minerals in the iron oxide.

In addition, I note that Dr Küzma initially concluded that the iron oxide “had not originated from the Kigali area where the Excellent Mining loading facility is located”. At the trial, she qualified this statement to say that the iron oxide had not originated only from the two mine sites in Nyaruvumu and Gituntu. I also note that the samples used for her report were only taken on 25 and 27 May 2021 by Alex Stewart, a few years after the thefts had occurred. However, there is no evidence that this has affected the accuracy of her report.

Having considered the experts’ opinions on both the 3Ts minerals and the iron oxide, I am not convinced by Dr Petrone’s opinion and I find Dr Küzma’s explanation more plausible.

The paint on the drums’ lids

The lids of all the drums were painted with white alkyd paint after the tin concentrate was loaded into the drums and the drums were sealed. After the circumference of the drums’ lids were cut and the contents replaced, the lids were welded to the top of the drums leaving the welded holes and the Precintia clips that sealed the holes of the drums untouched. This gave the officials who inspected the drums the false impression that the drums were untampered with. The defendant contends that a thick layer of white alkyd paint that was used to cover the welding marks on the drums’ lids was the same paint that was originally used to paint the drums.

Laboratory analysis of paint samples

At MSC in Penang, Dr Petrone collected some paint flakes from around the rim of a drum’s lid. In addition, a section of a drum’s lid was also cut out to obtain a 10cm × 10cm sample of the paint layer. These two samples were sent to the laboratory of SOCOTEC UK Ltd (“Socotec”) for analysis.

During Dr Petrone’s inspection at Excellent Mining’s premises, he found cans of paint in an open yard which he claimed contained the paint used on the drum’s lids. This paint was “Email Brillante, Super Gloss”, manufactured by Ameki Color, a paint manufacturer based in Kigali, Rwanda. Dr Petrone claimed that he and Socotec’s laboratory tried to contact the paint manufacturer to no avail. Hence, he was unable to obtain a can of the original paint for comparative purposes. When Dr Petrone was conducting his investigation at Excellent Mining he could have collected some paint for Socotec’s analysis. He claimed that he did not do so as paint is a highly flammable substance so he could not bring the can of paint with him when he took a flight back. This is a lame excuse as it is not necessary to have the whole can of paint for analysis. Some small quantity of white paint would be sufficient for analysis just as he had collected some white paint from the lid of the drum at MSC in Penang for analysis.

Socotec issued its report on 15 March 2019 (the “Socotec Report”). The report states the summary of its findings as follows:

A plain reading of the above findings indicates that, while there was a very strong correlation between the two samples of paint, the samples were ultimately not a complete match. This was mainly due to the fact that the samples were contaminated. Hence, it appears that the Socotec Report is inconclusive.

However, the plaintiff’s expert, Dr Küzma, and the defendant’s expert, Dr Petrone, arrived at diametrically opposed interpretations of the findings in the Socotec Report. Dr Küzma opined that the two white paints were not the same while Dr Petrone opined otherwise.

Dr Küzma’s interpretation was premised on the Socotec’s Fourier Transform Infrared Spectroscopy (FTIR) analysis. According to Socotec, this was used “to examine the top layer of the paint flake sample in order to determine the resin identity.” I reproduce the resin identification results (the “Resin ID”) below:

The first graph shows the Resin ID of the paint flakes at the rim of a drum’s lid, and the second graph is a sample of paint from the centre of a drum’s lid. The last graph is the overlaying of the two graphs for comparison.

According to Dr Küzma, the FTIR analysis basically gives the “fingerprint” of the compound, as FTIR results are unique for each compound. Hence, for the two paint samples to match, all the peaks in the two respective graphs must be identical. However, this was not the case here because there were numerous differences in the peaks of the two graphs. In her expert report, she stated as follows:

At the trial, she enumerated the many differences between the two graphs:

Besides comparing the graphs, Dr Küzma also compared the chemical composition of the two samples. I set out her testimony here in detail:

As stated above, she came to the conclusion that the two white alkyd paints were different.

Dr Petrone’s opinion was that the differences between the two graphs were not sufficiently significant to displace his conclusion that the paint samples were a match. In his expert report, he stated as follows:

At the trial, Dr Petrone sought to undermine Dr Küzma’s opinion, stating that her explanation was incorrect:

In my view, Dr Petrone’s testimony showed that he was overly eager to disprove Dr Küzma’s opinion and this again reveals his biased opinion to support the defendant.

Firstly, he said that the fingerprints of the two samples are “the same” and that Socotec “confirmed” this. Yet, Socotec’s report stated clearly that they did not identify the two samples as a complete match (see [116] above). In fact, this was consistent with the conclusion in his expert report on this issue:

Evidently, Dr Petrone was content to simply proceed on the basis of Socotec’s finding that there was a “very strong correlation” between the two paint samples, to conclude that the two samples are from the same paint.

Secondly, Dr Petrone said that he had listened to Dr Küzma’s evidence in court. Yet, his short description of her evidence clearly paled in comparison to the long and detailed explanation that she gave at the trial (see [119]–[120] above). In fact, Dr Küzma was even ready to elaborate on more differences between the two graphs.

In any case, it is apparent that Dr Küzma’s findings cohere with that of the Socotec Report. I find the opinion of Dr Küzma more convincing and reasonable when she opined that the two paint samples do not match, despite their strong correlation. On the weight of evidence, I find in favour of the plaintiff on the issue of the source of the paints.

Dr Petrone assumed that the same group of people must have applied this paint after the thefts were committed. Yet, Dr Petrone confirmed that the “Email Brillante, Super Gloss” paint is a commercially available one. That must clearly mean that other people in Rwanda or even the neighbouring State of Tanzania would have been able to buy and use this paint at the material time. No evidence was adduced to show that Excellent Mining was the only one using this paint then. In the circumstances, even if the defendant could show that the two paint samples were an exact match, this would not have assisted its case.

Photographs of the drums

Dr Petrone claims that the thick layer of white alkyd paint which was used to cover up the welding marks was present since the drums left Excellent Mining’s premises to the time they arrived in Penang. From this, he concludes that the drums’ lids had been cut open and rewelded, and the paint applied, by the time the drums left Excellent Mining’s premises. Therefore, the thefts had occurred there.

In his expert report, Dr Petrone explained the above in detail and referred to a photograph of the drums at Excellent Mining’s premises and another in Penang:

Dr Petrone’s theory was that the swap of the drums’ contents had to take place inside Excellent Mining’s premises. His expert opinion was that the thickness of the paint applied to the drums in both photographs was the same. Hence, the two photographs above show the state of the drums after the theft had occurred.

In my view, the fact that the photographs above show the same thickness of paint is inconclusive. It could also suggest that the thief was very adept at concealing the welding marks. I wish to reiterate that the parties agree that the swap took place after the tin concentrate was loaded into the drums and the holes were welded and sealed.

In court Dr Petrone said his theory was premised on the fact that Excellent Mining’s staff was complicit in the thefts. This was because he acknowledged that thefts of this magnitude could not have gone unnoticed at Excellent Mining’s premises. At the trial, he conceded that, if Excellent Mining’s staff was not complicit in the thefts, his theory that the thefts occurred at Excellent Mining’s premises would not hold. Yet, as I have analysed above, there was no evidence to suggest that Excellent Mining’s staff was involved. To begin with, the level of security there was the highest compared to the other stages of the Transit Period (see [53]–[69] above). Moreover, there was no reason to impugn the findings of the NPPA Report (see [90]–[97] above). In the circumstances, Dr Petrone’s theory cannot stand.

Conclusion on the paint on the drums’ lids

Having considered the evidence in relation to the paint on the drums’ lids, I find that they do not show that the thefts had occurred at Excellent Mining’s premises.

Dr Petrone’s prejudiced and biased investigation

I shall now comment on Dr Petrone’s investigation into this case. Unfortunately, his investigation is unsatisfactory, superficial and substandard.

I begin with his investigation in Rwanda. This is Dr Petrone’s account of what he did in Rwanda:

Dr Petrone went to Rwanda for only two days and he spent an aggregate of only five to six hours carrying out the investigation. Out of this duration, he spent three hours at Excellent Mining’s premises, during which he spoke to Mr Sindikubwadbo and Mr Emmanuel Uwimana, a representative from Excellent Mining who assisted informally with the translations. Dr Petrone spent the remaining two hours at the Bonded Warehouse and the RIB. I note that he omitted many important tasks when he conducted his investigation:

(a) He did not speak with any representative or MFO from the RMB, despite knowing of the RMB’s function and being aware that there was an MFO stationed to monitor the drums of tin concentrate at Excellent Mining’s premises.

(b) At the trial, he accepted that Excellent Mining had security guards, but he claimed that he did not see any when he was there. His testimony therefore suggests that he did not speak to any security guards or ask the people at Excellent Mining’s premises whether there were security guards.

(c) He did not speak to any representative from ITSCI to obtain any information regarding the thefts or to find out from ITSCI if there were any similar cases of organised crime of such magnitude occurring in Rwanda.

(d) He did not speak to any representatives from the RRA or Alex Stewart to obtain any information regarding similar thefts of this magnitude.

(e) He did not go to the areas surrounding Excellent Mining’s premises to question neighbouring residents of any suspicious activities or to look for CCTV footage of the vicinity.

(f) He did not look for any cutting equipment that could have been used to cut the drums’ lids. In this regard, he confirmed that he did not find any welding equipment at Excellent Mining’s premises.

(g) He did not take any paint samples from Excellent Mining’s premises.

(h) He did not open the can of “Email Brillante, Super Gloss” paint he found at Excellent Mining’s premises to verify its contents. At the trial, he conceded that because of this omission, it was possible that the can of paint may not have contained white paint. There is nothing in his report which indicates that he questioned Excellent Mining’s personnel to confirm that this paint was used at Excellent Mining’s premises.

(i) He did not speak to any of the drivers from Bolloré Logistics who delivered the tin concentrate from Kigali to Dar es Salaam.

(j) He did not travel to Dar es Salam to carry out further investigations at the Inland Container Depot or the port. Mr Wheeler gave evidence that “graft in Tanzania is well documented, almost on a regular basis”, and a “quick scrutiny of news reports from local newspapers” showed that “over 10 per cent of goods are under peril of theft”. Yet, Dr Petrone did not know that the Dar es Salaam port was deemed as having a high risk for criminal activities at the time of his investigation.

(k) In respect of the NPPA Report and its findings, he claimed that he met RIB personnel yet he testified that he did not speak with them about their investigation when he was in Kigali.

(l) He did not arrange for the samples of the iron oxide that was swapped for the tin concentrate to be tested for their composition and origin.

(m) He did not take samples of iron oxide at Excellent Mining’s premises so that they could be tested for composition and origin as well.

(n) Although his theory that the thefts occurred at Excellent Mining’s premises is premised on the presence of the 3Ts minerals at that location (see [106] above), he did not take samples at Excellent Mining’s premises to test for the presence of these 3Ts minerals.

Ironically, despite the numerous omissions in his own investigation set out above, Dr Petrone accused the NPPA and the RIB of having done a lackadaisical and perfunctory investigation just from his mere review of the NPPA Report (see [90]–[97] above).

Moreover, Dr Petrone’s case at the trial was that Excellent Mining was complicit in the thefts. However, he omitted to mention this highly pertinent point in his expert report. He also did not mention in his affidavit of evidence-in-chief that Excellent Mining was complicit in the thefts. Such a drastic and late shift in position weighs heavily against his credibility as a witness.

In the circumstances, I find that Dr Petrone harboured a prejudiced belief and had a preconceived judgment that the thefts took place at Excellent Mining from the very start of his investigation. He testified as follows:

Having a predetermined outcome in his mind, his investigation therefore lacked objectivity, and this resulted in the numerous glaring omissions in his investigation above. I have therefore treated Dr Petrone’s evidence with utmost caution in coming to my decision.

The plaintiff has proven its case on a balance of probabilities

It is undisputed that the burden of proof is on the plaintiff to prove its case on a balance of probabilities. I am satisfied that the plaintiff has discharged its burden of proving that the thefts took place during the Transit Period and thus the Policy covers the thefts of the tin concentrate, for the following reasons:

(a) With regard to the level of security at various points, I find that security was at its weakest when the 40ft containers were transported from the Bonded Warehouse in Kigali to the port in Dar es Salaam. In contrast, security was at its highest at Excellent Mining’s premises. Hence, the weakest link in the whole journey must have been the former (see [52]–[89]).

(b) With regard to the NPPA Report, there is no credible evidence for me to doubt its findings that the thefts did not occur at Excellent Mining’s premises. I therefore accord weight to these findings in favour of the plaintiff (see [90]–[97]).

(c) With regard to the presence of the 3Ts minerals in the swapped iron oxide, I am unable to accept Dr Petrone’s theory that this suggests that the thefts took place at Excellent Mining’s premises (see [98]–[112]).

(d) With regard to the paint on the drums’ lids, I find that the laboratory analysis of the paint samples and the photographs of the drums do not show that the thefts had occurred in Excellent Mining’s premises (see [113]–[131]).

Conclusion

In view of my findings above, it is highly unlikely that the thefts had occurred at Excellent Mining’s premises. Rather, the thefts had most likely occurred during the journey from the Bonded Warehouse in Kigali to the port in Dar es Salaam. I therefore hold that the plaintiff has discharged its burden of proving, on a balance of probabilities, that the thefts had occurred during the Transit Period. Hence, I allow the plaintiff’s claim for the undisputed sum of US$1,154,508.94.