Wilex Cocoa & Coconut Products Ltd v Electric Power Corporation [2010] WSSC 107 (22 October 2010)

IN THE SUPREME COURT OF SAMOA
HELD AT MULINUU

BETWEEN:

WILEX COCOA & COCONUT PRODUCTS LTD
a duly incorporated company having its registered office in Apia
Plaintiff

AND:

ELECTRIC POWER CORPORATION
a statutory corporation duly established under the electric Power Act 1972
and continued under the Electric Power Corporation Act 1980
Defendant

Presiding Judge: Justice Slicer

Counsel: P Fepuleai and A Roma for plaintiff
M Ring Q.C, A Hooker and D Clarke for defendant

Hearing: 10-13, 15-20 August 2010
Judgement: 22 October 2010

REASONS OF THE COURT

1. On 1 June 2003, a fire destroyed the plaintiff's factory at Moata'a. The fire started in the mid-afternoon and, despite the best efforts of the fire service, had consumed most of the stock, interior fittings and structure by nightfall.
   
2. In July 2004, the plaintiff ("Wilex") commenced proceedings against the defendant ("EPC") to recover its loss caused by the fire, an action based on claims of negligence and breach of statutory duty. It relied on the statute to define the duty of care and its standard. The Electric Power Corporation Act 1980 ("The Act") sets out the powers and functions of EPC. The Act, section 15 relevantly states:
   

The functions of the Corporation are:

(a) To generate, transmit, distribute and sell electricity;

(b) To control and provide and maintain, on, over or under any land, generating and storage works, dams, roads, water races, penstocks, fuel storage facilities, electric lines, transformer stations or any other plant or equipment used in the generation, transmission or distribution of electricity;

(c) To promote, organise, co-ordinate, continue, develop and maintain the generation, transmission or distribution of electricity;

......

(g) To erect works and install all necessary machinery and plant for the purpose of generating electricity;

(h) To enter any contract for the sale, lease or purchase of property in relation to the generation, transmission, distribution or use of electricity...

3. The Corporation is given wide powers of inspection and entry; The Act ss34, 35. Those powers extend to its own installations 'being used for the generation, transmission, distribution...of electricity'; The Act s34 (4). Those powers create corresponding duties.
   
4. Here the parties agree that the duties and obligations of the Corporation are akin to those required by the common law and the principles governing the tort of negligence. It is not necessary to consider the statutory duty separate from negligence.
   
5. At trial, Wilex did not pursue its claim based on statute since the remedy claimed through tort was equally open and depended on the same particulars. Wilex relied on a failure of equipment operated and maintained by EPC and necessary for the supply of electricity to the factory. It maintained that a fault in a transformer caused a high voltage pulse to enter into the cable wiring inside the factory. Secondary contentions involved the possibilities that the fault had occurred at the adjacent power pole or at least there had been a failure in the 'cut off' fuses.
   
6. A separate cause of action concerned a claimed failure by EPC to promptly disconnect a high voltage connection, preventing fire-fighters from properly accessing the building in time to prevent significant or total damage.
   
7. EPC conducted its own investigation into the cause of the fire. It claimed that the origin of the fire was in the wiring connection of an air-conditioner to the power source and the cause, an improper repair to that wiring.
   
8. The plaintiff claimed that a reasonable assessment of loss exceeded ST$10 million. The defendant conceded loss but sought a far smaller figure in any award. It will be necessary to assess an award of damages irrespective of the findings on liability.
   
9. This is a causation case.
   

Legal Principles

10. The plaintiff is required to prove on the balance of probabilities that the defendant's negligence was the cause of loss. A plaintiff must first establish that an act or omission did cause the harm (McGhee v National Coal Board [1972] UKHL 7; [1973] 1 WLR 1 (HL)). In some instances there may be a sequence of events culminating in the harm, but the question remains whether there was causation in fact, of a character which the law, as a further matter of policy, will recognise (Deloitte Haskins & Sells v National Mutual Life Nominees (1991) 3 NZBLC 102, 259 (CA)).
    
11. Causation remains a question of fact. In law it differs from scientific notions of causations because 'questions of cause and consequence are not the same for law as for philosophy and science' (National Insurance Co. of New Zealand Ltd v Espagne [1961] HCA 15; (1961) 105 CLR 569, Windeyer J at 591). "In philosophy and science, the concept of causation has been developed in the context of explaining conditions and occurrences. In law, on the other hand, problems of causations arise in the context of ascertaining or apportioning legal responsibility for a given occurrence (March v Stramare Pty Ltd [1991] HCA 12; (1990-1991) 171 CLR 506, Mason J at 509). There remains a tension, as here, when scientists provide opinions based on science on matters where the ultimate legal question is cause and effect.
    
12. That question is made more complex in cases involving causation and the measure of damages (see: Overseas Tankship (UK) Ltd v Miller Steamship Co. Pty (No2) – 'The Wagon Mound') which provide a foreseeability test; contributory negligence (Alford v Magee [1952] HCA 3; 1952 85 CLR 437; last opportunity or real cause) and novus actus interveniens (Dorset Yacht Club Co. v Home Office [1970] AC 1030).
    
13. In this case, the question whether the fire was a consequence of EPC's equipment is a straight question fact to be determined 'by applying commonsense to the facts of each case' (Stapley v Gypsum Mines Ltd [1953] UKHL 4; [1953] AC 663).
    
14. The courts have used differing terms to describe the necessary act or event which brought about the injury or damage such as 'effective cause', 'real and efficient cause', 'direct cause', 'proximate cause' and the like. There is judicial debate about the difference between a 'but for' test and the term 'sine qua non' (see generally: March v Stramare (supra), McHugh J at 525-534). Whether one starts with a 'but for' question and then looks for a policy (remoteness) limitation, or starts further along the line with a 'real and effective cause' question, looking back to 'but for' aspects and then onward to policy (remoteness), one will often end up at the same point. This is the law not logic (see generally: The Law of Torts in New Zealand (4th ed.), Todd 21.1, 21.2).
    
15. Here the parties and experts permit the court to apply a simpler question namely; has the plaintiff proved on the balance of probabilities that EPC operations or equipment brought about the fire which destroyed the plaintiff's property?
    

The Factory

16. The factory was a substantial building 1,285m2 in area set some 15 meters from the main road from Apia. It was set on a concrete slab and its external structure built of concrete blocks. The southern portion of the building comprised a second or mezzanine storey with a wooden exterior. The roof was metal and seen from the road, had two gables; one higher on the eastern side to provide for the mezzanine floor. There were two metal roller-doors; one facing the road and the second on the eastern side close to the southern corner. There was a 'fall back' generator on the western wall, not in use but which had a connecting cable running over an internal ceiling and separately joined to the main switchboard. The mains cable was threaded through metal trusses above the ceiling from the connection point to the main switchboard. The mains cable connected to the external power lines from the southern wall over an internal ceiling to the same master switchboard. The master switchboard was fixed to a wall between the packaging and cold rooms. There was a storeroom which shared a partitioning wall on the southern side of those rooms. The air-conditioner unit (said by EPC to be the cause of the fire) was fixed to the eastern wall of the storeroom. A copy of the floor plan of the factory site is annexed to these reasons for judgment.
    

EPC Equipment and Power Connection

17. Four power supply lines connected the internal factory cable to the power pole on the south-eastern corner of the factory. The fourth was the neutral wiring. They are joined at the lower arms of the pole to wires carrying the lower voltage current of 400V which also carry the current for ordinary households with standard electrical appliances. Given that the equipment used by the factory required greater current or volume known as 3-phase power, it was necessary to use more than the normal connecting wires than those supplying an ordinary household.
    
18. High voltage transmission was also carried along lines utilising the same telegraph poles. Some 400 meters west of the factory was a transformer. Simply put a transformer is a device which acts as a reducer of voltage. High voltage carried by special lines is reduced before connecting to lower voltage circuits. Here Wilex claims that a fault in the transformer allowed high voltage currents to transfer or jump into the carrier lines to the pole serving the factory and then along the connecting lines into the main Wilex cable. The higher current exceeded the safe designed volume and voltage of that cable which warmed. It is this warming throughout the length of the cable which in turn, through contact or radiation raised the temperature of a flammable substance or object, touching or near the cable beyond its ignition point which in turn spread to other objects.
    
19. The wires were fitted with circuit breakers designed to disengage if an excessive current passed through them. There was a suggestion that at least one of these devices failed, itself a failure causing greater harm to the Wilex system.
    
20. Wilex relies on the timing of the various stages of the fire, the observations of phenomena by independent witnesses to establish both cause and origin of the fire and together with the observed location of the initial smoke and flames to provide the basis and strength of opinion evidence supportive of its cause. EPC relies on the evidence of times of the reporting of the fire, the sequence of events as seen by the fire-fighters and a post-event investigation and analysis to meet that evidence and in turn to establish that the origin of the fire was the air-conditioner connection.
    

Onus of Proof

21. The onus of proof lies with the plaintiff to establish liability on the balance of probabilities. It must first establish the causative act or event and if successful, to show that such acts or omissions were negligent. Here the defendant is not required to establish that its scenario or evidentiary basis for its claimed cause is more probable then those advanced by the plaintiff. The plaintiff must show that the 'pulse' explanation is more probable than not. Here the defendant says that in the course of negating the plaintiff's case it is able, also to show that its own causative event is more probable.
    
22. A similar exercise is required in the establishing of significant facts.
    
23. In order to determine causation the court must first consider and decide identified primary facts.
    

Eye Witnesses

24. As one would expect eye-witnesses gave varying accounts of their observations depending on their respective vantages and the time of their presence at the scene.
    
25. Reverend Faafetai Milo put the time of his presence at about 3pm but he was still present when the fire-fighters arrived which suggests that his arrival at the scene was far later than 3pm. Maika Lemi saw sparking only at the lower of the cross-arms of the power poles. He thought smoke came from the building about 15-20 minutes before the arrival of the fire service.
    
26. Keneti Samoa claimed to have seen a woman fire-fighter. The evidence was that the only women in the service were volunteers and it is unlikely that a volunteer would be one of the first to enter the building. Keneti saw the fire-fighters using water to fight the fire early in the sequence of events.
    
27. Toenuti Fualau recalls hearing the noise and seeing the sparking after the second church bells i.e. after 3:30pm. She saw sparking only on the bottom wires. She estimated that the time from the sparking to the emission of smoke was about 10 minutes, an observation which supported EPC's case. Significantly she saw wires 'hanging down'; a matter consistent with the safety fuses having tripped.
    
28. Lesina Afa had made inconsistent statements to police, investigators and the court. Details of her evidence that she had spoken with a female fire-fighter who had told her that the live electricity had stopped the fighting of the fire and her observations inside the building are of little weight. She did confirm that the EPC crew arrived at about 4:10pm.
    
29. Siu Siu heard the loud bang shortly before 4pm which suggests that he heard the main switchboard explosion. Other evidence is consistent with the explosion that occurred after the arrival of the fire officers but little turns on that difference. Siu attempted to contact EPC but received no answer. The explanation for that failure was that the ordinary number of EPC during office hours was 22261 whilst the emergency number of 65400 and it is likely that Siu called the former.
    
30. The witnesses are generally consistent and their accounts are accepted as honestly made. But the details are a product of reconstruction and affected by the lapse of time.
    

Time of Events

31. Witnesses gave varying times of observations of smoke, flames and arcing, and different versions of noise such as crackling and a shot or explosion. Some of the observations differed in detail or sequence. Such is natural and within everyday experience. The court accepts variance, and will test accounts through records and documentation made close to the event and acceptance that memory can be affected by time. It is conscious that in many cases difference of accounts may be a product of honest reconstruction, self-interest and the passage of time.
    
32. The fire occurred on a Sunday after 3pm. The Reverend Faafetai Milo said that he saw 'blue sparking' flashing along EPC cables towards Wilex at about 3pm. He claimed to have seen two sections of sparking – one between the Penaia Brothers pole (the transformer) to the Wilex pole and a secondary arcing between that pole and the building. But he also claimed to have seen smoke immediately after the arcing. His observations do not accord with the opinion evidence given on behalf of either party. The sparking or pulse would not have produced immediate smoke. But if the time of his observation is later then there is consistency. Likewise with the evidence of Maika Lemi, a welder who heard a gunshot sound and saw sparking soon after 3pm. He later saw black smoke coming from the factory. Keneti Samoa put the 'gunshot' sound at about 3:20pm and the emission of black smoke soon after the second church bell of 3:30pm. Toenuti Fualau also fixed her time of observation of sparking by reference to the second church bell, and the smoke some 10 minutes later.
    
33. It is impossible to determine precisely the time of ignition. There is no method of determining the time of any 'pulse' event since on the scientific evidence any fire required ignition of a secondary event.
    
34. EPC contends that it is likely that any 'arcing' or 'sparking' was a product of that secondary event i.e. fire damaging the electrical system or its own form of short circuiting.
    
35. The 'gunshot' noise is likely to be the sound of an explosion in the main switchboard, an event which antedated the Wilex's ignition time.
    
36. Eddie Wilson, Wilex's managing director was contacted some short time before 4pm and immediately went to the scene. On his way, he telephoned EPC as verified by his cell-phone records at 4:01pm. He had been contacted by Lesina Afa, whose evidence will be later separately considered, shortly before 4pm. She in turn had been contacted by Kelese.
    
37. The fire service operator telephoned EPC at 3:57pm, confirmed by a log record. Lafaele Lefata, an EPC operator received that call and one made about 5 minutes earlier by an unknown man. That evidence differed from the time of 3:30pm recorded in a memo prepared some three weeks later. If it was Kelese who was the male caller to EPC, it suggests that the first call to EPC was between 3:40 and 3:50pm.
    
38. The fire service unit arrived at the scene at 4pm. The EPC team of Samasoni and Tokuma arrived between 4 and 4:15pm and disconnected the remaining fuses of the Wilex wires and later cut the supply cables. Their evidence will be later considered in relation to the question of delayed response.
    
39. Given that it is impossible to precisely determine the time between any event ('pulse' or 'wiring') and ignition, the court is able to estimate that ignition occurred between 3:15 and 3:30pm. It is possible to estimate that there were manifestations of that ignition and the spread of fire by 3:30 or 3:40pm. The original manifestation could have been arcing before smoke but it may be the reverse.
    
40. The above, coupled with the opinion evidence based on the laws of physics permit the court to make the finding that ignition, not prime cause, occurred before 3:30pm.
    

TIME	EVENT
3:00 to 3:30pm	ignition
3:15 to 3:30pm	phenomena possible but unlikely
3:30pm	manifestations, smoke, crackling and arcing likely
3:30 to 3:45pm	first phone call to EPC possibly by Kelese or another unnamed person
3:54pm	first notice to fire service
3:57pm	fire service notifies EPC
4:01pm	Wilson notifies EPC
4:00 to 4:10pm	EPC vehicle arrives
4:05 to 4:10pm	EPC disconnect 2-fuses

Phenomena

41. Witnesses gave differing accounts of the form and colour of the sparking and the accompanying noises of crackling and explosion. The court accepts the general tenor of those accounts and the integrity of the witnesses who gave evidence of their occurrence. It accepts that:
    

(1) There was arcing. Whether it was confined to the lines between the factory and the adjacent pole or extended to the lines between the transformer and the 'Wilex pole' is immaterial.

(2) On any version of the observable evidence or the opinions of those versed in electrical mechanics the arcing or jumping could have been a product or consequence of the inflammation, not a certain manifestation of its cause.

(3) It is possible that the arcing between the wires was a product of heat radiation or swaying produced by magnetic repulsion. It was not necessarily evidence of a 'spike' or 'pulse'.

(4) There were differences between the 'crackling' and the 'gunshot' noise. The former is consistent with natural or consequential arcing (Point 3). The latter with an explosive event touching the main switchboard.

(5) Evidence concerning the damage to the cable extending to the outside 'reserve' generator suggests an explosive switchboard event later in the sequence than would be possible if the initial pulse or surge had caused the damage.

(6) Evidence of dense black smoke, as observed by a number of witnesses supports the evidence of Kelderman and Oliver that it was the smoke itself which was a conduit for electrical terminal connection enlivening the generator switch.

(7) The evidence of prior sparking or flashing unconnected with an outbreak of fire is not an infrequent manifestation.

42. The above analysis shows that the phenomena evidence does not establish, of itself, that the arcing or gunshot evidence occurred before ignition. Arcing can be a product or consequence of the fire not associated with the cause.
    

Evidence of Scene Witnesses

43. One or more witnesses contacted the fire service. In the meantime, Lesina Afa, Wilson's stepmother and a resident of a fale at the northern end of the factory had been told of the fire shortly before her phone call to Wilson. She was alerted to the blaze and went to the eastern roller-door opening it to gain access to the interior. She said that she entered the production room and saw no fire or smoke in that area. Neither did she notice any smoke from the area of the meter box or master switchboard but did see smoke coming from the front of the building. She remained at the site when Wilson and his office supervisor George Mueller arrived.
    
44. Lesina acted as a caretaker for the site. Her evidence differed significantly from her accounts given to investigators shortly after the fire. In cross-examination she impressed as a partisan witness. The court accepts her general evidence but not its specific details. Her detailed description of observations within the building were those of reconstruction of an event long ago.
    
45. The same may be said of Wilson. He lost his most valuable asset in the fire. His personal interests and those of his company aided the process of reconstruction of the events. He was adamant in matters readily disproved by objective evidence and details of his account contradicted by independent witnesses. Again his general account of events is accepted but details, unless corroborated by others or objective evidence, are given little weight. His credibility on issues relating to quantum was weakened through cross-examination and witnesses called by the defendant. That in turn impacts on the accuracy of his account of what he saw upon entering the building.
    
46. Wilson had entered the building after the arrival of fire officers. There were many events happening at the same time and doubtless the scene was 'busy'. According to Luff, a fire investigator who interviewed Wilson shortly after the event:
    

...from memory, he entered the building with his maintenance manager, Mr George Mueller and also at the time was a female fire officer and there's sort of two versions one was at the whole area, the whole building was filled with smoke and there was another that there was only a layer of smoke on the ceiling but that Mr Wilson noticed that the upstairs office area...access in that he tried to get up there to remove some equipment, computer but it was restrained by fire service...

Q: So as far as what Mr Wilson told you what was the state of the smoke and fire in the production area when he entered in there?

A: "there was no fire and from memory the smoke was in the higher level or in the ceiling and it was making out of the front gable area."

Q: And did he tell you anything about having seen the state of the main switch board?

A: "that was no evidence of fire in or around the switch board when he went in."

47. He gave a different account to Marnix Kelderman, an experienced investigator shortly after the fire. Kelderman claimed that Wilson had told him that when he first arrived at the scene he had seen smoke coming from the front left gable of the building.
    
48. A further matter which impinges on Wilson's credibility was the delivery in June 2010 to Maanaima Taua of Lesina's statement (exhibit "P2.7"). Taua did not open the document but gave it to his CEO.
    

Evidence of Fire Officers

49. The fire service arrived at the scene within minutes of the call out. Fire officers immediately began to fight the fire from the front of the building specifically, at the western section near the front roller-door. In accordance with procedures they did not wait for the disconnection of electricity (Taua Ex Book 1, paragraph 1.4 to paragraph 8). Other officers entered the building through the eastern roller-door. Wilson was directed to leave the interior because of the risk of injury. Those officers were not concerned with any continued power outage.
    
50. The fire-fighters ran out of water and whilst further supplies were ordered the main assault on the fire was delayed. The fire spread out of control. It was after the officers had arrived that the switchboard exploded exacerbating both the fire and the destruction.
    

EPC Officers

51. EPC maintained a second emergency crew. The duty officer received a telephone call reporting the fire soon after 3:30pm, probably about 3:45pm. The emergency crew were notified. A second call was made from the fire service at 3:57pm. By that time the emergency crew were on their way. The senior officer Samasoni lived at Sogi which is some 5 minutes distance from the scene of the fire. He went to the home of his partner Tokuma Noaese and waited a brief time to collect and assemble his equipment. They arrived shortly after the fire service, probably shortly before 4:10pm. They found that two of the safety fuses had tripped. The scientific evidence was that the tripping of those fuses would stop the flow of electric current along the wires connecting to the factory junction. Tokuma immediately climbed the ladder. He could not see any flames inside the building from his vantage, an observation consistent with the opinion of Kelderman as to the course of the fire.
    
52. It is possible to work back from two established times. The fire service was notified at 3:54pm and the area supply cut off through ABS 34 at 4:30pm. The evidence of EPC workers and records was that the area supply was isolated approximately 15-20 minutes after the arrival of the emergency crew because of concerns for the integrity of the high voltage lines.
    

The Transformer

53. The transformer was affixed to a power pole some 400 meters west of the factory. It governed the operation of the current generated at the power station. The supply lines carried three lines conducting the current. The top carried the high voltage current of 22kV. That current was reduced to 400V for 3-phase operations and 1-phase power at 230V (Penaia Brothers, the plant nearest the transformer and Wilex were users of 3-phase power). The transformer was a sealed unit and was removed and inspected following the fire. It was found to be undamaged with no signs of arcing or scorch damage. The examination revealed no moisture or contaminates inside the unit, nor any degradation of the oil within the sealed unit.
    

Other Users

54. There was no evidence that domestic users within the immediate area experienced damage to household appliances at the time of the fire. There was no evidence that Penaia Brothers, users of 3-phase power, reported any damage to their plant and equipment.
    

INVESTIGATION

55. The fire, its cause and damage, was investigated by five separate persons or statutory bodies. The fire service and EPC conducted their own review of the incident and respective responses. Gary Luff was retained by the plaintiff to identify the cause of the fire whilst the defendant retained Marnix Kelderman, an experience fire investigator to inspect the site and damaged equipment and report his findings to it. Kelderman later received technical assistance from Ross Oliver, an electrical engineer who also visited the site and examined the damaged equipment.
    
56. Lawrence Kamm, a distinguished and experienced electrical engineer retained by the plaintiff did not visit the site until trial and relied on the evidence collected by others in formulating his opinion.
    

Fire Service and EPC

57. The records of the fire service help fix a timeframe for the course of the fire and fire officers provided independent accounts of their observations at the scene. They will be considered in relation of the action based on delay. The EPC provided its records and technical data to Oliver and many of those records were tendered at trial. A missing document was said by the plaintiff to show a cover up but, as will be seen later in these reasons dealing with the alternate cause of action its absence is of little import.
    
58. Fire cause and origin is a specialised discipline (Taulaga v S V Mackenzie & Company Limited [2010] WSSC, unrep. 2 August).
    
59. Luff, who was called as a witness by the defendant at trial found nothing which suggested that there had been a defect in the EPC equipment or method in its provision of power.
    
60. Kelderman has impressive credentials. He has a Bachelor of Science degree with a triple major in Botany, Zoology and Cellular and Molecular Biology and a Masters of Science degree. Between 1991 and 2001, he was employed as a forensic scientist and fire investigator by the New Zealand Department of Scientific and Industrial Research and since that date has operated his own company conducting fire investigations on behalf of New Zealand police and fire services and other insurance and overseas agencies. He has been a past President and Vice President of the New Zealand Forensic Science Society. He has attended and investigated some 2,000 fire scenes.
    
61. Kelderman attended and examined the fire scene from 16 June to 23 June 2003. He interviewed many of the primary witnesses who gave evidence at trial. He said that Wilson had told him that when he first arrived at the scene he saw smoke coming from the top of the building's front gable before entering the factory through the eastern door.
    
62. The witness was using his scientific knowledge, training and experience. His opinion and conclusions will be later considered. His findings are summarised in the following terms:
    

A (1) He examined the general patterns of burning and determined the general direction of the fire and specific directions in each localised area. A plan showing the various directions is annexed to these reasons for judgment.

(2) The patterns indicated that the fire had impinged on the office area at a later stage of the developing fire before rolling downwards to the area below.

(3) The fire had burnt into the laboratory through its northern side and continued south to the front of the building. There were no indications of arcing to the electrical wiring in that area.

(4) Examination of documents and stacked pallets in the 'records room' revealed that the fire had burnt across from the northern side.

(5) Fire damage in a 'walk in chiller' was consistent with a secondary fire, an observation confirmed by the re-ignition of a fire in that location requiring the fire service on the following day.

(6) Fire to the western front of the interior had entered through the large northern interior entrance.

(7) Pieces of cable in the production area had arced. The arcing had occurred while they were 'live' or 'energised' indicating that the main switchboard had not yet been damaged or itself 'arced' until the fire had developed.

(8) Severe damage to the cold room container could be explained by 'combustible gases produced during the breakdown of the polystyrene insulation under heat.'

(9) The main switchboard in the packaging room had been severely damaged. It contained a severely damaged generator changeover switch, which was significant to an opinion formed by Oliver. There was considerable evidence of arcing in the electrical components.

(10) There was no evidence of 'arcing' in the remains of the mains cable which had been moved to a different part of the floor. This indicated to Kelderman that the electricity was already disrupted when the cable was damaged. He would have expected to see melted copper conductors and damaged steel armouring had the cable been live.

(11) Fire had penetrated the western side of the packaging room coming from the southern wall.

(12) The cooler storeroom was used for the storing of packaged and wrapped product including unprocessed cocoa. That room was constantly maintained at a low temperature by an air-conditioner located above one of two doors on the eastern wall. Kelderman assumed the only door in use was the northern (closest to the rear of the building) and the southern door sealed off. Wilson disputed this assumption. He concluded from burn patterns below that southern door that fire had travelled from within the storeroom and moved outwards across the door lintel. Linoleum had burnt away under the air-conditioner whilst its opposite side had no such burning. Remains of the packaging debris under the southern door reinforced his opinion.

(13) The main fire had proceeded from the area of the cooled storeroom into the southern portion of the building.

B. Kelderman examined the exterior of the building and the northern sections and found nothing to alter his findings that the origin had been in the storeroom.

C. He then turned his attention to specific objects and items of equipment. He had found no 'arcing' in the mains cable. He found no damage to the supply cables to the gable, but had with wiring within the production area. Examination of the cable connecting the outside generator showed that it had been severed. The end connected to the 'change over switch showed electrical arcing' and the severed portion (previously disconnected from the generator) showed no arcing. The cable had been severed by a fallen object from above, a result of the fire. The 'arced' section indicated that it was 'live' at some stage of the fire. The explosion of the main switchboard would have destroyed the switch allowing plasma to flow into the cable causing the arcing.

D. The investigator next looked for counter-indications; the state of the EPC equipment and found no evidence of external cause. He viewed photographs taken from television footage when the fire was at its height. The nature of the fire and its location was consistent with his view that the fire had started towards the middle of the building. He was aware of the claim that the fire was a result of a fault in the EPC mains electricity supply. The intact gable at the point of indication showed that the fire could not have commenced at this point.

E. Examination of the debris produced objects significant to the formation of an opinion. They included:

(1) sections of the mains cable with no discernable arcing;

(2) the severed generator cable with only the connected end showing sparking;

(3) a metal table adjacent to the storeroom;

(4) the air-conditioner unit with attached flex; and

(5) segments of flex associated with (4).

63. The air-conditioner unit was found lying on the flooring on the eastern side of the wall to the storeroom. It had been buried by remains of the concrete block wall and timber framing and lay on top of burnt wood panelling. It had not been disturbed. Examination showed it to have been subject to intense heat. Perforations to the metal suggested that they were caused by degradation, rust and corrosion. The flex attached to the unit was one meter long and itself attached to another piece of flex. The remains showed that the attachment was made by the removal of the insulation, the entwinement of the wires and, presumably, recovered with tape. The two recovered segments had been twisted together in an identical manner.
    
64. Kelderman was unable to locate a plug socket.
    
65. Gary Luff, a New Zealand fire investigator appeared, under subpoena, as a defence witness. On 7 June 2003, he came to Samoa to investigate the scene at the request of the plaintiff. He did not confer with the defendant's counsel before giving evidence.
    
66. Luff conducted an investigation. He spoke with Wilson who told him that when he, Wilson, entered the premises the whole building was filled with smoke which went up to the ceiling.
    
67. Luff inspected the site. He found no evidence of damage to the transformer or the Wilex pole and none to the isolators or safety fuses. He found no evidence of electrical activity in the mains cable which, on his understanding was suspended across steel trusses two metres above the ceiling. He had told Wilson that he (Luff) had ruled out the mains cable as the vehicle for starting the fire. He concentrated on the main switchboard because of the damage, and found significant electrical activity consistent with the fault occurring 'downstream' i.e. from sources inside the factory rather than from the mains cable. The switchboard had its own 3-phase outlet to the disconnected generator and as at June 2003, Luff believed that the observed electrical activity had something to do with that connection, a suspicion which was consistent with Oliver's inspection of the same cable. He saw no evidence of a power surge or pulse into the main switchboard which he would have expected had there been one.
    
68. Luff located and examined the air-conditioner although he did not move it from its resting place. He did not see the flex which was buried or hidden below the unit. He found no signs of electrical activity with that unit although he had made a thorough investigation of that area. He sought material from EPC which was not provided. He conferred with Ian Alexander, a New Zealand expert in electrical engineering when he returned from Samoa. On 13 July 2003, Alexander replied seeking some suggestion and advice. That ended Luff's involvement in the matter until the subpoena.
    

Air Conditioner

69. The plaintiff contests the integrity of the finding pointing out that the site had been seriously compromised by the activity of fire-fighters and the clean up operations.
    
70. There was evidence that the unit was operating at the time of the fire. It was kept in operation at all times to protect the stored products. The plant itself was not in operation on the Sunday. It is unlikely that any other electrical appliance or equipment was active at the time of the fire.
    
71. Aiono Sam Aiono, a respected and experienced electrician gave evidence for the plaintiff and testified that the commercial unit was standard using a standard plug socket connected to an ordinary power outlet. He had serviced the factory's unit and had never noticed any improper or illegal wiring which he would have rectified. In cross-examination he was unable to recall when he had last serviced the particular unit, but believed that he would visit the plant at 3-4 month intervals.
    

OPINION EVIDENCE

72. The parties had considered but excluded other possible causes for the fire except the 'power surge' and 'air conditioner' events.
    
73. During the trial the respective expert witnesses conferred. They agreed that the above theories or analyses were the only plausible explanations for the cause of the fire.
    
74. The competing explanations can be stated as:
    

(1) The Plaintiff.

The origin of the fire was at a point below or adjacent to the 20 metre length of the mains cable between the southern gable and the main switchboard. The initiating event was either: an accidental physical connection between EPC's 22kV and 400V supply allowing a 22kV voltage into the Wilex cable.

OR

A defect in the transformer (probably moisture contamination) caused the 22kV current to enter the 400V line and into the Wilex system.

The secondary event was the ensuing pulse which caused the insulation between the two of the phased wirings to uniformly degrade and a resulting overheating under the resumed 400V load. That heat ignited nearby combustible material.

(2) The Defendant.

The improper splicing of the wiring in the connecting cord of the air-conditioner 'faulted' and the heating melted the insulation igniting the other combustible material which spread to adjoining areas, or itself started to burn allowing molten plastic to drop out flammable material.

75. Central to the competing contentions is identification of the place of origin of the fire.
    
76. The plaintiff retained Lawrence Kamm to provide an opinion of whether, and if so how, a failure of equipment or the delivery of service may have been responsible for the fire.
    
77. Kamm has Masters and Bachelor degrees in Electrical Engineering. He holds a Californian Licensed Electrical Engineering Certification, and since 1985 has served as an expert witness in cases involving electrical injuries, electrical fires or industrial accidents and analyzing their causes.
    
78. Kamm did not visit the site until trial. He did not examine any of the equipment. He reviewed the various reports, the documents and answers provided by the defendant, police and witness statements, photographs and other material later tendered at trial. He attended the hearing and followed the course of the plaintiff's case.
    
79. In his initial report of June 2003, he accepted as propositions:
    

(1) On a Sunday there was a small electrical load on the main switchboard.

(2) There were no visible defects in the Wilex wiring.

(3) There was no evidence of any other possible source of fire.

(4) The fire brigade believed that heavy smoke came from the location of the main switchboard.

(5) The cause of the fire was electrical because of (4).

(6) Fire-fighters withheld the use of water until EPC cut off the electricity.

(7) There was no evidence of a source of the fire on the Wilex side of the main switchboard.

(8) The delay of several hours' response by EPC in disconnecting power was responsible for the delay by the fire brigade in attacking the fire.

(9) There was evidence of arcing on the EPC conductors leading to Wilex.

(10) Mueller, an engineer employed at Wilex had stated that there were no poor connections within the Wilex system.

80. From those assumptions he concluded that 'sparking is evidence of such insulation failures...reports of sparking indicate that EPC failed to maintain proper insulation of their wires.'
    
81. Six of the ten assumptions were incorrect. One (5) was correct for the wrong reason, and a second (10) later proved to be incorrect.
    
82. In a second report of 21 July 2010, he gave as his opinion that:
    

"...the sparking observed by the witnesses suggests that 20,000 volts became momentarily connected to 400 volt wire connected to Wilex and was sufficient to perforate its insulation and establish an arc of sufficient power to ignite the insulation and thence the structure of the factory. The voltage could be wire to wire or wire to ground."

83. Following his arrival in Samoa he examined new evidence not previously made available and provided a proof of evidence dated 16 August (exhibit "P3") which included:
    

"It is my opinion that sparking observed by the witnesses suggest that the 20,000 volts became momentarily connected to Wilex and was sufficient to damage its insulation and the resulting heat was sufficient to ignite some of the fuel and thence the structure of the factory."

"My theory was that a high voltage pulse from EPC triggered overheating on the mains cable and that it was this heat that caused the fire."

84. He was not inclined to agree with the defence argument about the 'twisted wire' cause although he agreed that it was plausible. He had interviewed Sam Aiono, the electrician who had assured him that the unit was in good condition and the power wire had no splices, and believing Aiono, excluded the unit or its connection as a possible cause.
    
85. At trial, he again rejected the 'twisted wire' version since no one would be so stupid as to repair electrical equipment in such a way.
    
86. In his evidence at trial, Kamm advanced a further possible cause namely that defects in the transformer, either through inadequate maintenance or the intrusion of moisture could have accounted for the 'surge' or spike of the 20kV current entering the 400V lines. He was firm in his opinion that a momentary surge would not necessarily cause arcing within the Wilex cable. The mechanism he suggested was that the surge would heat the cable which would itself generate sufficient heat to ignite a nearby or touching inflammable object (paper, rag, debris and the like) in the ceiling. He conceded that such an object could be between 1-2 meters away from the heat source.
    
87. The problem with this 'secondary' mechanism is that the evidence shows that the cable was threaded through a steel framework in the gable and ran along metal framework, two meters above the ceiling before its connection with the switchboard.
    
88. Kamm's reliance on the evidence of sparking was valid but depended on the form and location of the sparking. Only one witness, the Reverend Faafetai Milo claimed to have seen sparking on the 22kV lines between the transformer and the Wilex pole and he could easily have mistaken them for the 400V lines. There are other mechanisms for sparking, as described by the witnesses which, will be later considered.
    
89. No defects or moisture contamination had been found in the transformer. In his résumé (First Report Document, Exhibit Book 172) Kamm listed his successful cases in which he had been involved, one of which was that:
    

"I rebutted the 'expert' allegation that an overloaded transformer started a fire by examining the transformer and showing that its insulation was not damaged."

90. Three phase power derives from the same source as a consumer 1-phase supply namely, the 400V lines. The difference is that 3-phase power is derived from 2-phase conductors, a domestic 230V user from only one with a neutral conductor. No other consumer complained there had been damage to domestic appliances.
    
91. The defendant relied upon the evidence of Ross Oliver and Kelderman to meet the causative event advanced by Kamm and detail an alternate explanation for the origin of the fire. Oliver holds a Bachelor's degree in Engineering (Electrical and Electronic) and is a member of the New Zealand Institute of Professional Engineers. He has acted as an expert in many New Zealand cases and international arbitrations. He has prepared electrical designs for power generating stations and assisted in their construction and operations in eight countries in Asia, Africa, South America and the Pacific including Samoa. He has over 20 years of investigation and reporting on supply or equipment failures and fire incidents.
    
92. He was retained to examine the EPC plant and equipment, its procedures and records. He was asked to consider whether there was a fault within the EPC system or the Wilex factory. He conducted a rigorous examination of the EPC operation from its central generators to its point of delivery. In 1993, he had assisted in the design and construction of the Fale o le Fe'e hydro-plant and the Tanugamanono diesel station. He had a working knowledge of the Upolu grid, and problems, geographical and meteorological, specific to Samoa.
    
93. He examined every item of equipment, installation and transmission facility associated with this case. The transformer 3/2000 and its surge diverters and fuses connected to the unit were found to be intact. He looked, and where possible, tested the elements of equipment both external and internal to Wilex for evidence of insufficient maintenance or fault from an external event. He personally examined the transformer and all the service and transmission facilities connected to the factory and repeated his testing of the transformer in August 2010. He found no evidence of arcing or damage to EPC facilities.
    
94. Given that witnesses had observed 'sparking' he arranged on 12 August 2010 to view, from a cherry picker, the overhead lines between the Penaia Brothers transformer and the Wilex factory. He found that two 400/230V lines had damage to some of their strands. That damage was consistent with two of the low voltage lines touching. He also found some minor melting of aluminium conductors which had been caused by fire. The significance of those findings will be separately considered.
    
95. Oliver heard Kamm's testimony and reviewed it, before commenting on it during his own evidence. His salient points can be summarised as follows:
    

(1) Kamm made incorrect assumptions about the configuration of the transformer, relying on his own knowledge of a U.S. 2-phase transformer, not the 3-phase model used in New Zealand, the Pacific, the U.K. and Europe.

(2) Smaller transformers such as here require regular testing, not automatic reconditioning as Kamm claimed.

(3) The Samoan unit is specially sealed so that moisture cannot intrude unless the seal is disturbed.

(4) The seal in question had not been disturbed or opened since manufacture.

(5) Entry of moisture through 'breathing' or osmosis is prevented by a seal designed to withstand both changes in atmospheric pressure, operating temperatures, unlike the 'relay panel' model analogy used by Kamm.

(6) There can be catastrophic failure of oil filled transformers but they require long term degradation of oil, bad maintenance, manufacturing defect or gross overloading. Here there was no evidence of either a catastrophe or any of the above. The unit remained in operation until 16 June when it was removed for forensic examination. There had been no pulse.

(7) It is inconceivable that a 22kV surge could enter a domestic dwelling and cause no damage to household appliances.

(8) Any high voltage connection to a low voltage wire would cause permanent damage to the low voltage wiring of the transformer causing an immediate failure. This unit remained in operation for two weeks after the event.

(9) There was no arcing on the Wilex cable. Kamm claimed that a 22kV pulse travelling along the EPC lines would cause arcing between lines. Oliver said that if such were the case, the same pulse travelled through the Wilex cable 'between the exposed conductors and lugs where the mains cable joins the switchboard.' There was no damage to the conductors or the circuit breaker. This fact alone would exclude the possibility that a 22kV travelled down the mains cable.

(10) The claim that the mains cable had degraded in a semi-uniform manner along its full length has never been shown in practice or reported in the literature.

96. Oliver also said that from his experience in New Zealand, fire-fighters do not refrain from using hoses on low voltage equipment. They are advised not to direct water onto high voltage equipment but he, Oliver had never heard of any 'not fighting a fire until the electric power had been disconnected from a building.' Kelderman who has trained fire service personnel had never heard, over 20 years involvement, of the practice of waiting because the power had to be disconnected.
    
97. Kelderman also took issue with a number of Kamm's assumptions and conclusions which included:
    

(1) Smoke from a fire above the ceiling would move upwards to the unceilinged sections and be forced to areas where no smoke was seen and could not have entered the office area as described by Wilson.

(2) There was nothing apart from the steel trusses carrying the cable or its protected insulation which was capable of ignition.

(3) Fire in the ceiling would have caused an earlier collapse of the production room ceiling.

98. Oliver continued with his evidence supportive of the explanation that the fire was a product of the faulty air-conditioning unit.
    
99. He had examined the site with Kelderman and used site observations and retrieved objects in reconstructing the sequence of events. He initially concentrated on the main switchboard, and attempted to reconstruct the sequence of events preceding its destruction. His analysis was that it was the fire which damaged the components causing the explosion not that a pulse of electricity which overloaded the system. He examined the emergency generator connections. The generator itself had been disconnected but the cable to the main switchboard remained in place. The cable had been severed by a fall of debris or structure. There was arcing to the cable between its connection to the main switchboard but no arcing in the severed portion to the generator.
    
100. Oliver reasoned:
     

(1) The severance of the generator cable had been caused by an object falling from above or some physical event during the fire.

(2) Electricity was still flowing through the cable at severance; hence the arcing.

(3) Electricity was still flowing through the main switchboard at a time when the fire was well established.

(4) A pulse of 22kV through the mains cable ought to have damaged the main switchboard and/or at least tripped the internal fuses.

(5) There was no such damage or tripping since electricity still flowed into the generator cable (most likely after the destruction of the switchover device when the fire had been well established).

(6) Hence there was no pulse.

(7) Examination of other sub-circuit cables and switches showed post-fire damage.

(8) Absent human involvement it was likely that there was another electrical culprit – electrical – within the building.

101. The retrieval of the air-conditioner and flex provided a suspect. The unit had fallen and found partially covered by concrete blocks, concrete beam and roofing iron. It was clear to Kelderman and Oliver that it had not been disturbed which explains why the flex had not been examined by Luff. The flex consisted of three badly burned flexible copper cable cores, two broken close to the unit and the third intact for about a meter. The third core had been twisted to another piece of similar flex showing a 'makeshift' cable join.
     
102. Oliver's opinion that such a connection was illegal and dangerous and did not provide a reliable electrical connection.
     
103. Further inspection revealed several strands with a section of flex with another makeshift joint. The above items were photographed and the objects tendered at trial.
     
104. The literature (Ignition Handbook, 2003, Babrauskas), relevant regulatory provisions (NZ Electrical Regulations, Rs 20, 57) codes (National Electrical Code, L 3.7.2.2 to 3.7.2.8; The North American National Electrical Code 110.14) forbids the practice of splicing.
     
105. Oliver summarised the physical mechanisms of why the join is dangerous and can cause heat build up as:
     

"...when a small electric current flows through the loose joint it eventually creates excessive temperatures that are high enough to ignite combustible materials such as the plastic sheath of the flex. We know from witness statements that the Wilex air conditioner was on prior to the fire, so there would be electric current flowing in the flex to the air conditioner at that time."

106. He summarised his findings as:
     

"A poorly made joint, once it begins to loosen, will start to overheat. Overheating at the joint will lead to degradation of the plastic insulation around the joint and also degradation of the plastic insulation around the flex. Over time, the degraded plastic eventually becomes combustible. The heat generated at the wiring joint can eventually be sufficient to cause ignition of the plastic sheath. Most plastic sheaths on electrical flexes are not flame retardant and once they catch fire they continue to burn. Burning plastic often melts as it burns and the melting flaming plastic will drop from the flex and onto material below the flex. If this material is combustible, then the fire can take hold and become a structure fire, as was seen in the Wilex fire."

107. Oliver stated his ultimate opinion and conclusion as:
     

"Based on my inspection of the electricity supply set up supplying the consumers from the transformer opposite Apia Park, I can say that the equipment, arrangement and maintenance of this mains supply equipment was of a high quality. I could find no faults in its set up and equipment at all. I have seen no evidence that the supply of electricity by EPC was in any way defective or substandard. In my opinion the supply system and components were well built, using quality materials and well maintained at the time of the fire.

Based on my investigations, I am absolutely certain that there was no significant power surge or other fault that could have caused a fire within the Wilex building. On the contrary, the investigations I have undertaken have confirmed that in my opinion the fire in the Wilex building started as a result of an electrical fault within the Wilex building, quite likely relating to the poor wiring connection in the flex for the air conditioning unit near the storeroom."

108. Kamm had maintained his analysis and conclusion stating that the 'AC unit' (air-conditioner) explanation was tenuous and speculative. He rejected it, in part because he could not imagine anyone being so stupid as to affect such a join, and he accepted the evidence of the maintenance electrician.
     
109. The critique of Kamm was not his skill and grasp of theory but of the assumptions or selection of sections of the evidence used in the formulation of his hypothesis.
     
110. Kelderman had examined the whole of the complex. He had interviewed many of the witnesses including Wilson. He had conferred with Oliver on technical matters, studied the reports of EPC and the fire service. He had examined the air-conditioner flex and the generator cable. He supported Oliver in regarding the 'twisted' join as substandard and inherently dangerous.
     
111. He used his knowledge and experience to trace the course of the fire and, given the observations of the fire-fighters and others, and the latter stage of the fire as recorded on TV footage to determine the timing of the course or stages of the fire. He was a thorough and impressive witness. His evidence was not shaken in competent cross-examination. It is not necessary to recount the details of his specific findings set out in D17.1.
     
112. His ultimate opinion and conclusion will suffice namely that:
     

"In my opinion, the fire at Wilex Industrial Complex started in the eastern side of the cool storeroom, located towards the rear of the factory. The most likely cause of the fire was the sub-standard connection between two flexible wires supplying power for the air-conditioner that was running continuously to keep the storeroom of finished products cool. The connection of the two wires consisted of the copper conductors (wires) twisted together. There were no proper connectors in the join. This twisted joint would heat up. Over time the hot joint can pyrolyse the insulation and surrounding materials until they catch alight."

113. Luff reached no firm conclusions following his short investigation. He had formed a tentative opinion that the fire had started at or near the main switchboard.
     
114. He did not expect that the fire had been caused at or near the mains cable since, if a nearby fire had attacked the cable he would have expected to see evidence of burning or damages and he found none. The suspension of the cable over metal trusses and the two meter distance from the ceiling was a further relevant factor.
     
115. In his examination at trial, Luff was asked about the heat transfer possibility from the mains cable to a nearby inflammable object. He believed that it would not generate sufficient heat to bring about that result.
     

Sparking

116. Kamm had regarded the evidence of witnesses as supportive of the 'pulse' explanation.
     
117. There was no dispute that reliable witnesses had seen flashing and heard noises described as 'crackling' or akin to a shot.
     
118. Oliver and Kelderman accepted the credibility of the witnesses but not their nuance or detail. Oliver summarised his analysis of their collective evidence by saying:
     

"I have considered the evidence from the various eye witnesses who were in the vicinity leading up to and during the fire. In my experience, what those witnesses described are common observations by laypeople in relation to a fire of this type, and is consistent with my own examination of the interior of the factory and EPC's components outside the factory, and my conclusions about the cause of the fire."

119. Uili Taisia gave evidence that on Saturday, the day prior to the fire, he had seen flashes on the power lines near the Wilex factory. Such sparking, not dangerous, can occur in atmospheric conditions of high humidity. It is sometimes accompanied by crackling and is a natural phenomenon called a Corona Discharge. It occurs when the air surrounding an electrical component ionises i.e. breaks down and the air 'glows' usually a blue/white colour. Oliver had observed the same natural phenomena whilst walking, early in the morning of 14 August 2010, past the Samoan Supreme Court. Whilst it may appear to a layperson to indicate faulty electrical equipment it is but a transient, harmless natural occurrence.
     
120. The question of the evidence of the observed sparking on 1 June was whether it was evidence of a pulse predating the fire or a consequence. On any version the main switchboard exploded well after the fire had developed, not simultaneous with any pulse.
     
121. Oliver considered the observations of the witnesses and his own findings of the conductors and transmission wires made on 12 August 2010.
     
122. Electric current flowing through a power line creates a magnetic field which interacts with magnetic fields of adjacent wires. The strength of the field depends on the magnitude of the current. If a short circuit occurs downstream the magnetic fields alter creating a repulsion force causing the lines to move out. When the current ceases, they return. If the kinetic energy is sufficiently strong the returning wires might touch, creating a second short circuit which creates a flash and spark with a repeat of the process described above. That process can intensify. When the downstream circuit breakers trip, the current at the trip point reverses and makes the dynamics more complex.
     
123. Thus a fire causing multiple short circuits within the Wilex premises while the master switchboard was still energised could cause secondary shorting faults in several spans between the premises and the transformer. This would explain the 'flashing' on the 400V lines.
     
124. Oliver had examined the damage to the master switchboard and found that the plastic insulation had been damaged by fire creating a further short circuit fault at the point of damage. In his opinion:
     

"...these short circuit faults created an increased current in the overhead lines between the Wilex pole and the transformer pole. When the explosion occurred in the main switchboard, this caused a massive increase in the current as well. These current increases could well have caused the outer pairs of overhead lines on Span 1, 2 and 3 to swing. It appears to me that these lines on Span 1 and on Span 3 had swung and then touched together several times."

125. He added a further explanation for any sparking along the lines between the EPC pole and the Wilex factory that:
     

"...when the short circuit faults in the Wilex premises and, most likely, those caused in and by the main switchboard explosion, caused the pole fuses in the Wilex power pole to eventually blow, this certainly would have been accompanied by sparks and smoke coming from the pole fuses themselves. It is possible that these sparks and his smoke would have appeared to be travelled towards the factory from the pole to and an observer seeing this from a distance."

126. At any scene of an accident or dramatic event, witnesses see different occurrences and have differing impressions. Multiple events become mixed in memory or lack detail and produce differing perceptions. The witnesses saw flashing consistent with a series of short circuits occurring after damage by fire and at a later time, a result of the explosion in the master switchboard.
     
127. The summary of Oliver's opinion was that the observed sparking was confined to the 400V lines and his conclusion was:
     

"(1) 400V overhead lines – short circuit faults within the factory caused by the fire, and/or second short circuit faults and/or resulted in swaying and waves in Spans 1 & 3, as a result of which the outer pairs of lines touched causing arcing multiple times;

(2) 22kV overhead lines – possibly mistaken, and actually saw sparking on 400V overhead lines instead;

(3) Insulated lines between Wilex pole and factory – possibly mistaken and actually saw what they thought were sparks travelling towards factory when pole fuses blew..."

ALTERNATE CAUSES OF ACTION

128. The plaintiff pleaded, in the alternative, that the defendant's failure to promptly attend the scene of the fire and disconnect the power connection inhibited the fire-fighters from fully attending to the fire because of the risk of electrocution. The relevant pleading in the Statement of Claim was stated in paragraph 9 which states:
     

"(d) The Defendant's staff failed to turn off electricity to the Wilex factory on the day in question even after being informed of the fire because they assumed that the feeder breaker equipment would have been triggered by the "sparking" and cut off the electricity to the Wilex factory;

(e) Even after the Defendant's Managing Director personally informed the Defendant's duty staff at or about 4:01pm on Sunday, 1st June 2003, to disconnect the electricity supply feeder to the Wilex factory, the Defendant's staff did not arrive at the Wilex factory until after 4:30pm of that same day: by which the fire had been allowed the chance to take hold..."

129. A second discreet basis was that:
     

"(f) Furthermore, the pole mounted fuses which the Defendant had chosen, installed, and maintained failed to cut off the electricity supply to the Wilex factory after the sparking had taken place: electricity was still being supplied to the Wilex factory when the fire took hold..."

130. The latter pleading may be readily met. There was uncontested evidence that when the EPC team arrived at the scene they found that two of the fuses had tripped. They immediately disconnected the two remaining. The evidence is that the tripped fuses had effectively prevented any further flow of electric current. There was no failure or breach of duty.
     
131. The primary pleading was initially based on a belief that the emergency crew was engaged in another call out at Lepa Aleipata which caused the response delay. That belief was incorrect since EPC maintained a 'back up' crew. At trial, the issue was maintained on the basis that there had been inordinate delay between notification and response. The onus remains with the plaintiff to show inordinate or negligent delay.
     
132. The fire service call to EPC is recorded in the log as 1557 with Pumper One on scene at 1600. If Lafaele's memo is correct the EPC crew would have been at the scene before the fire-fighters. My finding is that the correct time sequence is:
     

TIME	EVENT
1601 hours	Wilson calls.
1600 hours	Pumper One at scene.
1557 hours	fire service contacts EPC.
1556 hours	Pumper One responds to the call.
1554 hours	fire reported to fire service by Ruta Elia.

133. Allowing for error, the latest time of the anonymous call to EPC was 1545 hours. That accords with the other evidence that it took the emergency crew 15-20 minutes to arrive at the scene soon after the fire service. One can fix their arrival time at between 1605 hours and 1610 hours.
     
134. Lafaele Iefata was EPC's fault service operator. He received two phone calls, 5 minutes apart. The first from an unknown male and the other from the fire service recorded at 3:57pm. Wilson made a call recorded at 4:01pm. It is possible that they were the two calls. At trial, Lafaele agreed that the log document was missing and he was relying on memory as the basis for his evidence. Kelese had advised Lesina Afa of the fire who immediately contacted Wilson putting Kelese's call shortly before her contact with Wilson. That evidence supports Kelese's evidence that he had first noticed the fire between 3:30 and 4pm. Lafaele dispatched the emergency crew upon receipt of the first call. It was likely that the first call was received shortly before 4pm.
     
135. The fire service report mentioned concern about the risk of conductive 3-phase electricity which might have involved the ability of fire-fighters to play water directly onto potentially live equipment, advice consistent with the evidence of Oliver.
     
136. The fire service log records the first notification of the Wilex fire at 1554 hours and the dispatch at 1556 hours.
     
137. Lafaele's memorandum records a call from an unnamed person at 3:35pm and a fire service operator advice at 3:40pm. At trial he conceded that he had written up the memorandum some days later, and the best he could do from memory was to say that he received the call at 3:30 to 4pm. Other documents relating to the fire were lost. There is discrepancy between the records of the fire service and EPC as is often the case when there is no electronic or mechanical record to verify times. I accept the memo to be inaccurate for two clear reasons namely;
     

The 3:35pm notation reads:

"Rang then at that time to Samasoni (standby) for some help since the Fault Service were at Lepa Aleipata for..."

the 3:40pm notation reads:

"3:40pm Fire Department rang for some help and it was close to that time and it was close to the time Samasongi and Tokuma were to arrive to turn off the power to the house."

138. The latter suggests that he had already dispatched the crew and believed they were already close to the scene. The former suggests that he had dispatched them at or soon thereafter before the 3:40pm call.
     
139. Wilson's evidence that the supply lines between the pole and the factory were not disconnected until 4:30pm is not accepted. He may have been confused by events. The area supply was disconnected at 4:30pm as verified by the objective records. The area disconnection was ordered because of the concern that intensity of the fire itself may have caused damaged to the high voltage lines adjacent to the factory.
     
140. There is another way to verify my conclusion. The supply area was isolated at 4:30pm because of safety concerns, a matter verified by record. That was done by Noaese Fuiatua together with Samasoni some 15-20 minutes after Samasoni's arrival, which would put the first presence of the EPC team at approximately 4:10pm.
     
141. There remains a conclusive answer to the plaintiff's contention. No fire officer claimed to have been inhibited or prevented from discharging their duty because of a fear of live electricity. No procedure or practice required inaction until the power supply was disconnected. The evidence demonstrated that they did their duty and were only precluded from a proactive response when the water supply was replenished. Lesina's claim that she was told to the contrary by a female fire officer is not accepted or even if correct was wrong. All of the female officers of the service were members of the 'volunteer' section and only called in at a secondary stage. Wilson's claim that he was told the same by a police woman is likewise not accepted.
     
142. The response time was not causative of the damage. Even if it were so there is no negligence on the party of the officers of the EPC.
     

CONCLUSION

143. The onus on the issue of causation remains with the plaintiff.
     
144. My findings and reasons stated in summary form are:
     

(1) The plaintiff has not established that the fire originated in the ceiling below and/or adjacent to the Wilex cable.

(2) Fire damage and debris was not consistent with a ceiling fire origin explanation.

(3) Arcing found in wiring within the factory especially to the severed generator cable shows that there was electricity supply to the main switchboard continued well after the fire had developed.

(4) There was no sign of damage to the mains cable and the suggestion of uniform degradation of that cable along its whole length not supported by scientific practice or experimentation.

(5) There was no evidence of damage to the EPC transformer or supply equipment or wiring. The August 2010 observations of Oliver were more consistent with a post-fire event.

(6) There was no evidence of arcing to the EPC power lines.

(7) Any pulse or spike causing damage to the Wilex cable or equipment would have affected the appliances of other users of electricity in the area.

(8) There was evidence that the fuses tripped as designed.

(9) The detailed evidence of Wilson of his observations inside the factory is not accepted. This finding is not to be interpreted as a finding of dishonesty.

(10) I prefer the evidence of Kelderman and Oliver to that of Kamm, since the opinion of Kamm depended on facts and assumptions not proved or found to be erroneous.

145. The plaintiff has not established on the balance of probabilities that equipment, operation or a pulse caused by EPC's generation or supply of electricity caused a fire or damage to the plaintiff's factory or equipments.
     
146. In relation to the second cause of actions my findings and conclusions are:
     

(1) The area supply was disconnected at 4:30pm.

(2) The EPC emergency team arrived at the scene of the fire shortly after the fire service. They arrived at approximately 4:10pm.

(3) There was no delay between notification to the crew and their arrival at the scene, other than that required for travel.

(4) The plaintiff has not established that the EPC operator was contacted at 3:30pm nor that he delayed the dispatch of the emergency crew. The only confirmed notice was that by the fire service at 3:57pm and Wilson at 4:01pm. At best, the unidentified male caller would not have made, in the context of the other evidence a call earlier than 3:45pm.

(5) Peleiupu Kelese did not repeat his claim of calling EPC at 3:30pm at trial.

(6) The fire-fighters did not cease their efforts in combating the fire. They did not cease their efforts because of a fear of live electricity. There is no practice in Samoa, New Zealand or other countries prohibiting the use of water by fire-fighters until the electricity source has been disconnected.

(7) The fire service was inhibited in its efforts because their water supply had been exhausted.

147. As to the ancillary contention of a failure of the EPC safety fuses, the court finds that two did function as designed and at that stage the flow of 3-phase electricity ceased. They were designed to trigger in response to a short circuit in the main switchboard.
     
148. The plaintiff has not established, on the balance of the probabilities that the EPC response caused additional damage to the plaintiff's property.
     
149. The defendant claimed a different and identified cause. It is not necessary in the determination of the outcome of the action to make findings concerning that cause. But it is appropriate to do so, especially given the seriousness of the event, as a guide to others and in deference to the work of the investigators.
     
150. My findings are:
     

(1) I accept the evidence given by Luff, Kelderman and Oliver.

(2) The origin of the fire was in or adjacent to the dividing wall of the storeroom.

(3) The primary cause was a defect in the wiring connecting an air-conditioning unit to a normal plug outlet. The defect was a repair or alteration to the wiring by the joining of two sections by a twisting of the wires to form a connection. There may have been more than one such connection. There was no suggestion that it had been made by Sam Aiono, the contract electrician.

(4) The form of connection was inherently dangerous and prohibited in most countries by regulation or codes of practice.

(5) The unit was in constant operation and was the only electrical appliance operating at the relevant time.

(6) The faulty connection generated heat which degraded the insulation either melting it or setting the coating on fire. The heat itself or the actual burning ignited other inflammable material.

(7) The unit was fixed onto the lintel above the unused door to the storeroom. The most likely flammable material was packaging stored below the unit.

(8) Electricity still flowed into the main switchboard during the early stages of the fire.

(9) The fire developed in the form as reconstructed by Kelderman. As it did so it caused further short circuits in other appliances.

(10) The fire impacted on the main switchboard causing it to explode. That catastrophic damage exacerbated the intensity of the fire and damage.

(11) The phenomena observed by witnesses was an effect or consequence of damage caused by fire not evidence of its cause.

Addendum

151. Counsel will note that no mention is made of the computer generated reconstruction of the course of the fire. The exhibit was displayed during the course of Oliver's evidence. It was objective in nature and of the highest professional standard. Oliver made use of the 'aid' to demonstrate his testimony. I did not review the exhibit until after I had completed my final draft of these reasons for judgment. I delayed my review because it is a 'beguiling' reconstruction and I preferred to rely on the primary evidence in reaching my conclusions. My review was solely to see if I had omitted any primary evidence in my analysis of the evidence. My decision was not through lack of respect for quality of the exhibit but because of its possible influence. I commend the author or maker of the 'aide'.
     
152. The action of the plaintiff is dismissed.
     

DAMAGES

153. The plaintiff claims the sum of $10,014,713.00 in damages. It abandoned its claim for loss of income. It is entitled to be placed in the same position it would have been but for the fire. Its claim will be assessed on the basis of replacement of its building, plant, equipment and stock and a separate claim for loss of interest. The court ought to assess damages irrespective of its finding on liability.
     

Replacement Cost of Building

154. The plaintiff relied on the calculations of Faanimo Warren, the Principal Land Valuer of the Ministry of Natural Resources and Environment. Her calculations resulted in a replacement cost of $2,657,000.00. The defendant contended that the replacement cost ought to be assessed as at March 2002. The figure is based on Wilex's valuation at a time when it proposed selling the building to a separate corporate entity albeit one controlled by the same family. The plaintiff's claim is to be preferred. It would be entitled to the cost of demolition and replacement of its premises. Wilex does not own the land which is held by customary title. Wilex had leased the land from the Wilson family and presumably could negotiate a fresh lease. Wilex constructed the building in 1994 and undertook extensions in 1998.
     
155. While the creditors might have sought to realize their securities and entered into possession arrangements might have been made, but for the fire, for continued occupation.
     
156. The plaintiff's claim under this head of damage ought to be upheld.
     
157. A separate claim was made for the cost of replacement of the concrete apron used for parking. At trial, the evidence showed damage to the concrete had occurred before the fire. No allowance is made for replacement.
     
158. The sum of $2,685,000 is assessed.
     

Plant and Equipment

159. Wilson prepared a schedule itemising the items and assessing their value. The inventory, supporting documents and reconstructions of records lost in the fire were supplied to Su'a Rimoni Ah Chong, an experienced charted accountant for assessment and calculation.
     
160. Su'a Rimoni was asked to provide a report on:
     

(1) the assessment of the value of lost plant and equipment; and

(2) value of lost stock.

161. His assessment for lost stock of $280,000 was accepted by the defendant. His figures for the plant and equipment were disputed.
     
162. The first challenge was whether the depreciation figure took into account that some of the plant installed in 1994 was new or second-hand. The second was whether a twenty or forty year depreciation period should be allowed. The third is whether certain items had been destroyed. The last was whether the court could have regard to the 'Lloveras quotation' concerned with the replacement of the cocoa processing plant.
     

Valuation and Method

163. The court rejects the submission that Su'a Rimoni acted as an advocate rather than an impartial expert as unwarranted and ungracious. The accountant was faced with a difficult task with little time for preparation. In advising Wilson that it was preferable to use a forty rather than a twenty year depreciation time was appropriate and accorded with accounting practice. The equipment was that of a previous technology, not based on planned obsolescence or modern and changing electronics.
     
164. The defendant's alternate assessment failed to take into account the effect of inflation or monetary fluctuation in determining the real cost of replacement.
     
165. A deduction will be made from the final figure assessed to take into account the fact that some of the equipment was second-hand and that the replacement value figure based on Wilson's estimates relied on should be slightly lessened.
     

Kava Processing Plant

166. The plant was recovered and repaired and has been moved to another site operated by the Wilson family or corporation. There was no evidence that it was damaged by the fire. The claim is disallowed.
     

Cocoa Processing Plant

167. The plaintiff obtained quotes from Lloveras, a Spanish company providing details of the cost of the components necessary for a full working plant. The report dated March 2004, tendered at trial, contained performance criteria and capacity and nine separate sections, running to fifteen pages, itemising forty-eight identified components necessary for the process. The defendant objected to the reception of the document since its maker was not called as a witness. The authenticity of the document was not put in question. The document was in the form of a business record and its contents clear. The best a cross-examiner could do would be to quibble with the amounts and have the witness agree that some other supplier might be prepared to give a slightly reduced price.
     
168. The court has received into evidence the document "P11.7". In so far as is necessary, it will exercise its power afforded by the Evidence Ordinance 1961, section 27 and declares that it was not reasonably practical to have called the maker of the document from Spain, a place beyond the seas to prove the documentation tending to establish a fact.
     
169. This is little prejudice to the defendant. Lloveras could have sent a record keeper to prove the document who would be unable to answer questions of specific costs and technical detail. But the document would have been admitted and only its weight criticised. Lloveras could have sent a senior officer able to answer detailed questions, a course which given the length and vagaries of the trial greatly increased the cost of attendance. There remains the obvious point of compliance by an involuntary witness. A subpoena issued from Samoa would be a difficult document to enforce in Spain especially since it could not guarantee a fixed return date.
     
170. The defendant ought to be well aware of the geographic isolation of Samoa and the difficulties of obtaining European technical advice or witnesses. It was ill-advised to make the objection.
     
171. The evidence is received and acted upon as establishing facts of its contents.
     
172. The quotation was for the amount of €2,684,084.
     

Balance of Equipment

173. The plaintiff accepts that the claims of $250,000 and $93,000 were included in the building replacement costs and will be deducted.
     
174. Wilson's evidence on the value of the equipment and the financial affairs of the company generally were the subject of criticism by the defence. On seven occasions he swore that all items on the schedule had been destroyed when quite clearly they had not. His representations to the banks and general dealings showed that his claims should be treated with caution. Su'a Rimoni was obliged to accept his (Wilson's) valuation at face-value in his formation of the claim.
     
175. The court does not accept Wilson as a credible witness on the issues of valuation. His evidence was inconsistent both internally within the trial and with previous statements. His answers given at trial were evasive and self-serving. But the court accepts that Wilex suffered significant loss.
     
176. The court will take into account the overvaluation of the costs as provided by Wilson. They will be reduced by 10% of their stated figure.
     

Interest

177. The plaintiff claimed interest incurred by the company and shareholders as a result of loss of return from the operations of the company. The defendant did not assail the integrity of Su'a Rimoni in his calculations.
     
178. Those calculations were:
     

ANZ	$676,488
DBS	$ 80,323
IFAC	$605,902
Total	$1,362,713

179. In 2008, the Wilson group entered into an agreement with ANZ Bank which resulted in the group paying $580,000 in full satisfaction of outstanding debts. The defendant claims that since the indebtedness of Wilex amounted to $831,000 it received a benefit of $251,000 which should be passed on to EPC. The contention is rejected.
     
180. EPC claims all the benefit without regard to the proportionate benefit received by the other entities or their creditors. It may be that the arrangement would not have been necessary had Wilex continued to trade. It had, after all just developed two near products before the fire.
     
181. Wilex was one of a number of entities within the Wilson group. The agreement with the ANZ Bank shows that the group as a whole or at least some of the components were indebted and presumably moneys paid as interest by shareholders related to the debts as a whole. The defendant claims there is no direct evidence of payments but the court infers from the bank records and the reconstruction of the accounts by Su'a Rimoni that moneys were paid. One half of the sum will be allowed for the DBS claim. Wilex is entitled to recover for interest paid in the sense recognised in Hungerfords v Walker [1989] HCA 8; (1988-9) 171 CLR 125. The interest component will be reduced to an amount of $1,322,552.
     

Summary

182. Many damage claims require a 'broad brush' approach (McElroy Miline v Commercial Electronics Ltd [1993] 1 NZLR 39 (CA)) and such will be taken here. No deductions will be made to the claims for building replacement, or stock but a minor one for interest. Specific deductions will be made from the plant and equipment and a further discount made for identified items with an overall discount for the age of the equipment and the accuracy of the Wilson valuations.
     
183. The assessments are:
     

Building Replacement	$2,685,000
Stock to be allowed in full	$ 280,000

Plant and Equipment was claimed in the sum of $5,787,183. Deductions will be made for:

Kava Plant - $ 40,000

Double claim for office repairs - $343,000

Total - $383,000

This results in a total of $5,404,183.

Two further deductions will be made for:

Depreciation adjustments - 10%

Allowance for overvaluation - 10%

This results in a further reduction of $1,080,826.

This final calculation becomes:

$5,404,183

Less $1,080,826

$4,323,357

This allowance for plant and equipment is $4,323,357.

Summary

Building Replacementt	$2,685,000
Plant and Equipment	$4,323,357
Lost Interest	$1,322,552
Stock	$ 280,000
Total	$8,610,909

184. The amount of $8,610,909 would be the appropriate award for damages.
     
185. Given the findings in relation to the 'delay' pleading it is not necessary to make any assessment of that portion of the assessed damages which ought to be awarded to the plaintiff.
     

ORDER

186. The action is dismissed and there will be judgment for the defendant.
     

(JUSTICE SLICER)