General Jones Pty Ltd v Wildridge & Sinclair Pty Ltd & Burnie Port Authority; Olympic General v Burnie Port Authority & Wildridge & Sinclair; Burnie Port Authority v Wildridge & Sinclair Pty Ltd & Olympic General - [1988] TASSC 38

Serial No 28/1988
List “A”

CITATION:General Jones Pty Ltd v Wildridge & Sinclair Pty Ltd & Burnie Port Authority;
Olympic General v Burnie Port Authority & Wildridge & Sinclair;
Burnie Port Authority v Wildridge & Sinclair Pty Ltd & Olympic General [1988] TASSC 38; A28/1988

PARTIES:  GENERAL JONES PTY LTD
  v
  WILDRIDGE & SINCLAIR PTY LTD

BURNIE PORT AUTHORITY
OLYMPIC GENERAL
v
BURNIE PORT AUTHORITY
WILDRIDGE & SINCLAIR PTY LTD
BURNIE PORT AUTHORITY
v
WILDRIDGE & SINCLAIR PTY LTD
OLYMPIC GENERAL

TITLE OF COURT:  SUPREME COURT OF TASMANIA
JURISDICTION:  ORIGINAL
FILE NO/S:  694/1981

2913/1981
2860A/1982

DELIVERED ON:  2 August 1988
DELIVERED AT:
HEARING DATE:
JUDGMENT OF:

Judgment Number: A28/1988
Number of paragraphs: 158

Serial No 28/1988
List "A"
File Nos 694/1981
2913/1981
2860A/1982

GENERAL JONES PTY LTD v WILDRIDGE & SINCLAIR PTY LTD
& BURNIE PORT AUTHORITY
OLYMPIC GENERAL v BURNIE PORT AUTHORITY
& WILDRIDGE & SINCLAIR
BURNIE PORT AUTHORITY v WILDRIDGE & SINCLAIR PTY LTD
& OLYMPIC GENERAL

REASONS FOR JUDGMENT NEASEY J

2 August 1988

TABLE OF CONTENTS

1         The Fire  1
2         The Proceedings  2
3         Preliminary issue – effect of failure to call certain witnesses  3
4         Principal elements involved in determining the cause of the fire  6
(a)      Stacking and location of the cartons of Isolite  6
(b)      Air movement  8
(c)      Combustion characteristics of Isolite  9
(d)      Occurrence of the fire  10
(e)      Burden of proof  12
(f)      The time element  14
(g)      The distance factor  15
(h)      The content of expert evidence  16

5         Determination as to the cause of the fire  22
6         Liability of Wildridge & Sinclair Pty Ltd  24
(a)      To General Jones Pty Ltd in Action No 694 of 1981  24
(b)      To Burnie Port Authority in Actions No 2860A of 1982 and 694 of 1981                27
(c)      To Olympic General Products Pty Ltd in Action No 2913 of 1981  29

7         Liability of Olympic General Products Pty Ltd  32
(a)      To Burnie Port Authority in Action No 2860A of 1982  32
(b)      To Wildridge & Sinclair Pty Ltd in Action  33

8         Liability of Burnie Port Authority No 694 of 1981  33
(1)      To General Jones Pty Ltd in Action No 694 of 1981  33
(a)      For escape of fire  33
(b)      In negligence  38
(2)      To Olympic General Products Pty Ltd in Action No 2913 of 1981  41

9 Summary of determinations as to liability 41

Serial No 28/1988
List "A"

File Nos 694/1981
2913/1981
286OA/1982

REASONS FOR JUDGMENT NEASEY J

2 August 1988

The Fire

At about 12.20pm on 20 December 1979, a fire broke out in the space between the ceiling and the roof of the Burnie Port Authority's cold store in the port area at Burnie. This cold store was a large rectangular building some 98 metres long and 28.8 metres wide. Part of it, referred to in the case as Stage 1, had been finished and was in operation. The fire occurred in the above–ceiling space in Stage 2, which was an extension of Stage 1. Stage 2 was then nearing the end of its construction. Stage 1 was divided by transverse partitions into three large rooms, numbered 1, 2 and 3 from the western end. (Directional references given throughout these reasons conform with those adopted for convenience during the hearing, in which "site north" is approximately 45 degrees west of true north – see exhibit BP 34, p2). Stage 2 was to consist of two rooms numbered 4 and 5, partitioned longitudinally. The fire erupted and spread very rapidly. It destroyed most of Stage 2, and a large part of Stage 1 up to and including part of room 1. Held in the cold store at the time under a contract with the Burnie Port Authority was a large quantity of frozen foods owned by General Jones Pty Ltd

The space between the ceiling and the roof in this building was called the "roof void", presumably because the roof and outer walls were not structurally attached to the refrigeration chamber itself, which was like a large oblong box enclosed by walls and floor and ceiling which were built with materials which have particularly good insulating qualities. These were panels in which the core consisted of expanded polystyrene ("EPS") between facing sheets of thin metal. The wall panels had aluminium on the exterior surface and steel sheet on the interior surface. The ceiling panels had aluminium on both sides. The panels were 250 millimetres thick for external walls, 200 millimetres for ceilings, and 150 millimetres for floor. The floor panels had a wearing concrete floor cast on top of them, and other supporting material underneath.

The walls of the cold store were supported by a light steel framework. The ceiling was supported by the walls, and by chains hanging from the roof. The outer structure of the building consisted of a series of steel portal frames which were set at approximately 8.3 metre centres and roofed over with asbestos cement sheeting. The above–ceiling shape of the portals was an inverted v The inner distance between the top of the ceiling and the underside of the apex was approximately three metres. At the eastern and western ends of the building there was a light gable construction, in which the exterior surface panels were again asbestos cement sheeting, which reached down a little below the tops of the walls of the cold store. All around the perimeter of the ceiling was a space approximately one metre wide, which meant that there was, as a design feature of the building, passage for a free flow of air from the outside right across the top of the ceiling. The drawings and photographs tendered in evidence amply illustrate the relevant construction details.

At the time of the fire seven men employed by Wildridge and Sinclair Pty Ltd were working in the roof void. They were erecting refrigerating pipes which were to carry the refrigerant, liquid ammonia, to the refrigerating units ("FDC's"), which were to be slung at appropriate intervals just under the ceiling. It is common ground that the seat of the fire was in a stack of cartons of insulation material consisting of EPS, the trade name of which was "Isolite", which was supplied to Wildridge and Sinclair Pty Ltd by Olympic General Products Pty Ltd The Burnie Port Authority and the three companies mentioned are the parties in the cases now before me.

The Proceedings

There are three actions. By No 6941981, General Jones Pty Ltd ("GJ") sues Wildridge and Sinclair Pty Ltd ("WS") and the Burnie Port Authority ("BPA") as first and second defendants, and those two defendants have joined Olympic General Products Pty Ltd ("OGP") as Third Party. The plaintiff alleges against WS that the fire was caused by welding operations carried out negligently within a few metres of a stack of approximately 30 cardboard cartons containing Isolite pipe insulation, which employees of WS had negligently stacked in the roof void above room 5, towards the south eastern corner of the building. The plaintiff claims against BPA that it was negligent and acted in breach of agreement in various respects in and about the carrying out of these welding operations by WS, and that it failed to use reasonable care to prevent the escape or spread of fire from Stage 2 to Stage 1 of the cold store where its goods were stored, causing their loss.

In this action, the first and second defendant each claims indemnity or contribution against OGP as Third Party. The first defendant, WS, alleges that the Isolite supplied by OGP was readily flammable and highly susceptible to ignite and catch fire, that OGP should have taken reasonable steps to draw these facts to the attention of WS, and negligently failed to give adequate warning thereof to them. The second defendant, BPA, claims against the third party on generally similar grounds.

In this action No 6941981, also, each of the defendants has served a notice upon the other claiming indemnity and/or contribution, based upon allegations against the other of the same general character as those made by the plaintiff against that other defendant.

In action No 29131981, the plaintiff, OGP, sues BPA as first defendant and WS as second defendant. The plaintiff's claim against BPA is based upon allegations that it acted negligently and in breach of its contract with the plaintiff by failing to take adequate safety precautions or to supervise adequately the use of welding equipment by employees of the second defendant, failing to clear combustible material from the area where the fire took place, failing to provide adequate fire fighting procedures and equipment, and the like. The plaintiff's claim against the second defendant, WS, is based upon alleged negligence by that defendant in its performance of the welding operations, thereby causing the fire.

Each defendant has served a notice alternatively seeking indemnity or contribution by the other.

The third action, No 2860A1982, is by BPA against WS as first defendant and OGP as second defendant. In its amended statement of claim, the plaintiff alleges that WS acted negligently and in breach of contract, by failing to install the pipe insulation in a proper, safe and workmanlike manner, and negligently storing and stacking the cartons of Isolite in the roof void, whereby the fire was caused to occur during the negligent carrying out of welding operations. Against the second defendant, OGP, the plaintiff alleges negligence in and concerning the supply of the Isolite insulation materials, by failing to test them adequately for flammability, failing to give adequate warnings and the like; the allegations being generally of the same character as those made against OGP by WS and BPA in action No 6941981. In 2860A there is also a counterclaim by the first defendant, WS, against the plaintiff BPA, for the value of the work done by WS up to the time of the fire.

Thus, causation of the fire, and in particular whether it was caused by negligent performance of welding operations near a stack of cartons of Isolite pipe insulation in the roof void of the Port Authority's cold store, is a common element in the principal issues involved in each of the three actions. As a consequence, it was ordered by Cosgrove J during interlocutory proceedings that questions of liability and of contribution between the various parties should be tried at the same time, in respect of all three actions and between all the contending parties. The following questions were ordered to be tried at the same time, and have been tried before me; each as a separate issue in the relevant action, before the trial of any question of quantum of damages:–

In Action No 694 of 1981–

(a)the question of liability of WS and BPA for the destruction by fire of the goods of GJ;
(b)the degree of contribution owed by WS and BPA to each other, and the degree of contribution owed by OGP to BPA and WS.

In Action No 2913 of 1981 –

The question of liability of BPA and WS to OGP, arising out of the destruction by fire of Stage 2 and part of Stage 1 of BPA's cold store at Burnie; and questions of contribution as between all three parties.

In Action No 2860A of 1982 –

The question of liability of WS and OGP to BPA for the destruction by fire of Stage 2 and part of Stage 1, and questions of contribution as between all three parties.

Preliminary issue – effect of failure to call certain witnesses

I shall deal first with an issue which may be regarded as preliminary. The hearing occupied approximately five weeks of court time, but most of that was taken up by technical evidence dealing with the kinds and characteristics of products emitted by the arc welding process, including in particular, distances over which those products can be thrown out from the welding point, and their ability to ignite materials of various kinds after emission.

It was common ground, or at least it was not contested, that some 20 to 30 cardboard cartons in which were contained pieces of Isolite pipe insulation, commonly called "lagging", were on the day before the fire hauled up into the roof void through the gap at the eastern end of the ceiling of Stage 2, and stacked in the roof void above room 5 not far from the eastern end. The cartons were stacked in an approximately rectangular heap. There was evidence, in substance not contested, that some minutes before the fire erupted, an employee of WS had completed some arc welding around the edges of a rectangular metal plate which was being joined to the underside of the flange of Portal No 12, which is the second last from west to east, and was performing that welding within a few metres of the stack of cartons which caught fire. The evidence as to the exact position in which the cartons were stacked, and the like, will be the subject of close examination later.

Of the seven employees of WS present in the roof void at the time the fire started, all had been doing work of various kinds on that morning connected with the extension of refrigeration pipes into Stage 2 from Stage 1. One would have thought it likely that all would have been called to give evidence by one party or another, and that thereby the full extent of available evidence concerning the occurrence of the fire would have been placed before the court as being relevant to the vital issue as to its cause. However, only two of those seven persons gave evidence. Those not called included the person who had done the welding, a Mr Ian Whyman. Of course, counsel for the various parties had their reasons, but the court was left without the evidence of a number of people whose testimony would seem to be relevant. It was submitted by the plaintiffs, and primarily by senior counsel for BPA, Mr Barnard, that I should draw inferences against the party whose only role was that of defendant, WS, arising out of the fact that that party did not call any of these witnesses.

I do not accept that submission. The relevant rule is stated in this way by Newton and Norris JJ in O'Donnell v Reichard [1975] VR 916, at 929:–

"(where) a party without explanation fails to call as a witness a person whom he might reasonably be expected to call, if that person's evidence would be favourable to him, then, although the jury may not treat as evidence what they may as a matter of speculation think that that person would have said if he had been called as a witness, nevertheless it is open to the jury to infer that that person's evidence would not have helped that party's case; if the jury draw that inference, then they may properly take it into account against the party in question for two purposes, namely: (a) in deciding whether to accept any particular evidence, which has in fact been given, either for or against that party, and which relates to a matter with respect to which the person not called as a witness could have spoken; and (b) in deciding whether to draw inferences of fact, which are open to them upon evidence which has been given, again in relation to matters with respect to which the person not called as a witness could have spoken."

The rule as so stated is derived from Jones v Dunkel (1959) 101 CLR 298 (see particularly per Kitto J at p308, and Menzies J at p312); The Insurance Commissioner v Joyce (1948) 77 CLR 39, per Rich J at p49, Dixon J, as he then was, at p61; May v O'Sullivan (1955) 92 CLR 654; and the other cases cited by Newton and Norris JJ in O'Donnell v Reichard (supra) at p929.

But the rule only applies where without explanation a party fails to give evidence or fails to call a witness when there is a reasonable expectation that he shall do so; which is to say, where the circumstances are such that if he does not, an adverse inference within proper limits may be drawn. The present is not such a case, in my view, nor do the three authorities cited by Mr Barnard in support of his argument persuade me that it is. In Jones v Dunkel (supra), the circumstances were such that there was certainly a clear expectation that the defendant driver would have been called if he had evidence which would have assisted the defendants' case. Earl v Castlemaine District Community Hospital [1974] VR 722 was another such case. It was a workers' compensation action in which a female plaintiff who was a waitress at a hospital sued for damages for injuries sustained when her finger was caught in a door handle. There was evidence that prior to the accident a problem with the handle of the door in question had been reported to the safety officer, who had new handles fitted some time after the accident. The safety officer was not called by the hospital to give relevant evidence about the door handle, which he above all people must have been able to give. Clearly there was room for an adverse inference to be drawn.

The third case cited, O'Donnell v Reichard (supra), was also of that kind. The plaintiff sued for personal injuries arising out of a car accident. She recovered damages which the defendant/appellant contended were excessive, on the ground that she was a malingerer and had failed to mitigate damages. The plaintiff had been treated by two senior psychiatrists and another resident doctor at the hospital, but she failed to call any of these doctors The trial judge did not direct the jury that an inference adverse to the plaintiff could be drawn in respect of this failure, and a Full Court of the Supreme Court of Victoria held that there had been a misdirection. One reason the trial judge stated for declining to give the requested direction was that the doctors could just as easily have been called by the defendant, but Gillard J stated the reasons why this was not so. His Honour said:–

"It might be pointed out that the power of the plaintiff to call the medical witnesses was much stronger than that of the defendant. The plaintiff was aware of the names of the medical practitioners who immediately attended her after the accident. She alone was in a position to authorize the hospital to give a report to her solicitor of the treatment she received in hospital and so obtain the names of medical witnesses, if necessary. Furthermore, having regard to the confidential relationship that would have existed between the plaintiff and the medical practitioners, she alone, or those acting for her, could have obtained precognition of the evidence that the doctors could give. Furthermore, if the defendant, despite the ignorance of what evidence each of such medical practitioners could give, elected to call any or all of them to give evidence, he would have been faced with the privilege that was open to the plaintiff to prevent their giving evidence under the provisions of s28 of the Evidence Act 1958. The position of the two parties was quite different in relation to these doctors, and it would have been quite unrealistic and unfair to expect the defendant to call the witnesses, and, accordingly, it placed some burden upon the plaintiff to explain why they were not called by her." (ibid, at p923.)

His Honour cited Earl v Castlemaine District Community Hospital (supra).

In the present case there was in my opinion no such reasonable expectation that the defendant, WS, should call Mr Whyman, or any of the others to the extent that an adverse inference should be drawn from its failure to call them. It was a fundamental part of the plaintiffs' cases that a building operation was carried out negligently in the roof void, and the fire was thereby caused. Presumably it was essential for the plaintiffs to call at least some evidence as to the circumstances in which the fire occurred, so as to provide material upon the basis of which an affirmative inference on that issue could be drawn on a balance of probabilities. The only witnesses who could give such evidence were the persons in the roof void at the time, who were all then employees of WS, as most were still. In the event, the plaintiffs collectively called only Mr Lillico and Mr Anderson, who, though employees of WS at the time of the fire, were not so at the time of trial. Those two witnesses were not able to give any evidence as to the time element between the completion of the welding operation and the commencement of the fire, which was an important element in proof of negligence. Presumably Mr Whyman, who, it was virtually common ground, had been the welder who had carried out the suspect operation, was the person most likely to be able to give such evidence. Over and above those considerations, though, was the highly relevant fact for this purpose that none of the parties was in doubt as to the substance of the account of the fire given by Mr Whyman, or any of the other employees in the roof void. It was proved at the trial that Mr Whyman had given evidence at the Inquest held into the fire in 1981 (his deposition being tendered), and it was a reasonable inference that all the other employees in the roof void (or all those then available who had anything useful to say) would have given evidence at the Inquest also. There was never any suggestion at the trial that there was any gap in the evidence given at that Inquiry.

The result must be that, in effect, all parties were in possession of the evidence given by all relevant witnesses at the Inquest, from which it is scarcely likely that they would or could in practice have departed substantially at the trial. Any of the plaintiffs could have tendered in evidence at the trial the depositions, or most of them, taken at the Inquest, by use of s81B of the Evidence Act 1910. Mr Lillico's deposition was put in that way. Therefore, an inference might just as readily be drawn that if evidence given at the Inquest would have assisted the plaintiffs' cases, as it might well have done, it would have been called by one of the plaintiffs as readily as by the defendant, WS. It is true that any relevant witness called by the plaintiffs, including Mr Whyman, could have been cross–examined by counsel for WS, but he could also have been cross–examined by other plaintiffs. In all the circumstances, I decline to draw an inference against any party in relation to the failure to call Mr Whyman or any of the other men who were in the roof void at the time of the fire.

Principal elements involved in determining the cause of the fire

I propose to discuss now these various factors

(a) Stacking and location of the cartons of Isolite

These cartons of Isolite measured 1205mm x 910mm x 470mm; ie, approximately 4 ft x 3 ft x 18 ins A piece of the cardboard of which the empty cartons were made was put in evidence. It appears to be about two millimetres thick, and to consist of two thin outer layers joined by some thin S–shaped material. That is to say, it is of cellular construction, with much air space between the two outer panels. Each piece of the enclosed Isolite (a piece was tendered in evidence) was a half cylinder, 90cms long x 31.5cms in diameter, with a piece cut out along its flat side, to accommodate the pipe which it is designed to insulate. Two of such pieces fit together around the pipe for each 90cm of length.

Three witnesses gave evidence that the cartons were hoisted from the ground into the eastern end of Stage 2 through the gap at the gable end of the ceiling. Mr Cooper, an engineer who was employed on the site by BPA as an assistant to its then acting Port Engineer, Mr Berwick, saw a number of such cartons, which he estimated to be between 20 and 30, being hauled up using a rope and pulley, on the afternoon before the day of the fire. One man was on the ground and another in the roof void receiving the cartons. Mr Cooper entered the roof void while that operation was going on, in order to speak with the foreman of the WS welding gang, Mr Ian Whyman. Mr Cooper said there was only one man receiving the cartons, and he was not stacking but throwing them to one side at the eastern end of the ceiling, just on the southern side of the centre one of the three vertical steel mullions (though he conceded his recollection was uncertain about this last point).

Mr Cooper's evidence in this respect differed from that of Mr Lillico, who was the WS employee receiving the cartons at ceiling level. Mr Lillico is a self–taught welder, then employed by WS. He gave evidence at the trial, nearly nine years after the fire and some seven years after the Coroner's Inquest. He said at the trial he had not been asked before giving evidence to refresh his memory of the evidence he gave at the Inquest. Understandably, his memory at the trial was hazy about some details, and he agreed that his Inquest evidence would be more accurate. I accept that. The full deposition of Mr Lillico's evidence at the Inquest was tendered before me, pursuant to s81B of the Evidence Act. Both at the trial and at the Inquest, Mr Lillico said that he had been receiving the cartons near the southern of the three mullions, and was handing each carton back to another man who was stacking them roughly as he received them. However, I prefer Mr Cooper's evidence to Mr Lillico's for accuracy about details, and where they differ on this point, I accept Mr Cooper. This means that the cartons must have been stacked at some later time, before the fire.

It was common ground at the trial that at the time of the fire there were two RSJ's mounted between portals 12 and 13 on the southern part of the eastern end of Stage 2. These two RSJ's ran east and west, parallel with the apex beam of the steel framing, and were equal in length to the distance between the portals (8.3m). They were suspended just above the top of the ceiling by two vertical steel beams welded to each end of the longitudinal pieces, and attached to the underside of the portal by a steel plate welded to each. Welded to the underside of each of the longitudinal pieces, towards the eastern end, were two steel cross channels, underneath which an FDC would have been suspended, by steel rods attached to the cross pieces and passed through the ceiling. The position of these steel members in the roof void after the fire is shown in photograph 5 of exhibit BP1. It was also common ground that the arc welding operation carried out by Mr Whyman on the morning of the fire, which is alleged to have been the cause of it, was carried out at the point where the connecting plate at the top of the vertical steel member holding up the western end of the northernmost of the two longitudinal steel beams was joined to the underside of portal 12.

Mr Lillico was questioned closely, both at the Inquest and the trial, as to the position and the manner in which the cartons of Isolite were stacked. That was a vital issue of fact in relation to the fire, in respect of whether the stack was close enough to the weld point so that hot particles emitted by the welding process could have reached the stack and ignited it. Mr Lillico's evidence on this matter varied somewhat, but in substance his best recollection, as given in evidence at the Inquest, was as follows. He received approximately between 20 and 30 cartons in the roof void. He was standing at the eastern gable end near the two longitudinal RSJ's used to support the FDC, and was passing the cartons back to another man who was stacking them just on the southern side of those RSJ's. I have already said that I prefer Mr Cooper's evidence on this latter point and that the cartons must have been stacked later. However, I see no reason not to accept at face value Mr Lillico's evidence as to how they were stacked.

Mr Lillico said in substance that the cartons were stacked in a rectangular pile, running east and west, about 10 feet wide by about 12–15 feet long, with the bottom layer on edge, and two top layers laid flat. The northern edge of the stack was about one foot or so south of the southern of the two RSJ's used to sling the FDC, there being room for a person to walk between the RSJ and the stack. The pile was stacked roughly. That was repeated by the witness several times during his evidence at the trial. He meant that the cartons were not perfectly aligned in a rectangular pile. A couple of cartons had been removed before the fire so that their contents could be used in the work. The eastern end of the stack was 10 feet or so from the eastern end of the ceiling.

I accept Mr Lillico's evidence, as I do that of all the witnesses, as being his and their best recollection honestly stated, but I emphasise that Mr Lillico's evidence about the size, shape and position of the stack of cartons is no more than his best memory of matters which would not have been particularly vivid to him at the time, and too high a value should not be placed on its likely accuracy. This is in contrast with his recollection of the fire itself, which would have made a much more striking impact upon his mind.

Mr Lillico's account of the position of the cartons in the three layers means that the height of the stack would have been approximately six feet. Since the roof above the cartons sloped from centre north to south, then the distance between the top of the pile and the roof would have been small. An indication of what that distance would probably have been is shown in Mr Hutchings' plan, No 7048–1 (attached to WS11). However, since the number of cartons is uncertain, varying between 20 and 30, it remains unclear whether the stack was a complete rectangular pile. Probably it was not. Nor is it clear from Mr Lillico's evidence, or the evidence overall, what the precise dimensions and shape of the stack were, except that it was "generally rectangular" or oblong. In his evidence at the trial, Mr Lillico said at one stage that the stack was "fairly square, about as high as a man and rough". At another point, he said "the bottom layer would be stacked with the narrower side down and then there were probably some on top of that on the flat, a couple of layers".

At the Inquest, when asked if he could remember how many boxes came up and were stacked, Mr Lillico said, "Not exactly. Roughly 20 or 30. … There was one layer standing upright and one or two layers on the flat on top of the upright layer." He was questioned more closely at the trial, when his recollection was less clear, than he was at the Inquest about the shape, location and dimensions of the stack; but he said at the Inquest that the eastern end of the stack was about six to eight feet from the eastern edge of the ceiling. At the trial, having been reminded that at the Inquest he had said that the cartons were stacked on the southern side of the southernmost suspended RSJ, he said that he had thought about that matter since being reminded, and was not really sure which side of the RSJ the stack was on. As already noted, he said that there were one or two layers on top of the bottom layer, but on another occasion he said he thought that was mostly the way they were. Overall, I conclude that Mr Lillico's estimates of distance and descriptions of the size, appearance and location of the stack or heap of cartons are very much rough estimates about matters which he had no particular reason to take note of before the fire, and has had to try to recollect or reconstruct in his mind afterwards.

The only other witness who was working in the roof void at the time of the fire was Mr Nigel Anderson. He was then a refrigeration mechanic employed by WS, and working in the above–ceiling area setting up valves in the refrigeration piping. He was working approximately under the centre line of the building above Stage 2 when he heard a shout of "fire", and looked towards the eastern end. He was about 15 metres from the stack of Isolite cartons. It was a regular stack. They were stacked neatly. Later, in cross–examination – the stack of boxes was "relatively neat for a size like that. Yes, they were sort of packed up fairly neatly, there might have been one or two loose boxes to one side, but yes generally very neat." He could not recall how much space there was between the top of the boxes and the roof, but there was clear headspace through there, and the pile was two boxes high but he could not recall the size of the stack. The stack was about six feet high. The boxes were stacked in from the edge of the ceiling, there would have been a clear space around them, there was a bit of a gap between the "eastern" end of the stack and the gap, but "in how big it was I – I really can't be accurate with it". The substance, then, of Mr Anderson's evidence about the stack of boxes, which I accept, was that it was a relatively neat stack, about six feet high, and there was some distance between its eastern edge and the edge of the ceiling, but he could not estimate that distance. He did not give any evidence about the dimensions of the stack or its location in relation to the suspended RSJ's.

(b) Air Movement

The question of air flow in the roof void at the time of the fire is a matter of importance, in an elevated area like this, some 8 metres above the ground, with the possibility of air flow all around due to the gap between the edge of the ceiling and the roof. It would be a matter of common knowledge that there would be likely at any given time in that area to be some movements of air; but that is as far as common knowledge could go. What the strength or direction of such air movement, if any, might have been at any given time would be a matter for evidence.

A certified extract from a meteorological report was tendered (BP4), showing that at 9.00 a.m. on 20 December 1979, there was a two knot wind speed from the south west, the air temperature was 13^o C and the weather fine. By 3.00pm the wind speed was two knots from the north west and the other conditions were unchanged. The engineer, Mr Berwick, said that when he arrived at the site of the fire at about 12.30pm, the wind was blowing in the direction of the hospital – which would make it a north westerly wind, and very dense black smoke from the fire was "heading towards the hospital". Photograph No 10 in exhibit BP1, which by appearance was taken when the fire was well advanced, shows the hospital and the dense smoke moving mainly vertically but to some extent towards the south east; but where smoke is seen emerging from the roof of the cold store, there is a definite movement towards the south east. I would say, therefore, it is probable that the wind speed and direction at the time of the fire had moved around so as to be approximately the same as at 3pm.

Dr MR Davis, who is head of the Department of Civil and Mechanical Engineering at the University of Tasmania, and whose specialty is unsteady aerodynamics within the field of aeronautics, was an expert witness called mainly as to other matters, but was asked about the possible movement of air within the roof void. The design of the roof and walls and ceiling of the cold store were described to him, and he was asked whether a two knot outside breeze would have a significant effect upon the flight of particles emitted by a welding process within the roof void. He replied that a two knot outside breeze, having regard to the particular construction of the building, would represent at the most in his opinion a quarter of a metre per second movement of air within the roof space. That would be an upper limit on wind speed within that space, and he thought it would be considerably less in practice. In his view such a wind speed would be likely to cause perhaps an additional two or three per cent variation in the horizontal distance spaces of the travel of such particles. I accept that evidence, and conclude that for practical purposes, air movement within the roof void, at least up to the point when ignition occurred in the cartons of Isolite, and the fire had been established, cannot be treated in any specific way as being relevant to resolution of the issues here. That does not mean, however, that a chance draught or movement of air may not in fact have fanned a piece of molten matter which came in contact with say a torn piece of cardboard inside a crevice, or the like. One cannot be specific about any such possibilities, but they cannot be eliminated from consideration.

(c) Combustion characteristics of Isolite

The chemical composition of Isolite, its ability to be ignited, and to burn once ignited, are all important matters in the case.

Dr BW Cherry, who is Associate Professor of Materials Engineering at Monash University, gave evidence about these matters. His principal specialty for the last 18 years or so has been polymer physics, related principally to the end use properties of solid polymers. The following description of Isolite is a summary taken from various parts of his evidence. Isolite is a commercial name under which a form of expanded polystyrene (EPS) is sold in various shapes and forms for insulation purposes connected with, amongst other things, the construction of cold stores. In the present case it was used as "lagging", which is an all round insulating cover for pipes carrying liquid ammonia used as a refrigerant in such buildings; and also in the form of floor, wall and ceiling panels of various thicknesses, ordinarily faced on either side by thin metal sheets. In the course of manufacturing Isolite of various building grades, substances are added for different purposes. Some are added in order to act as fire retardants. Styrene is a volatile liquid consisting of molecules of styrene, which can be polymerised. Polymerisation is the linking together of individual styrene molecules to form polystyrene, which is a hard plastic. Granules of polystyrene can be expanded, perhaps 20–fold, to form expanded polystyrene, which can be manufactured in various shapes and sizes. EPS is white and opaque, very light, and because of its close cell structure is an excellent thermal insulant.

When EPS is exposed to a heat source it softens and liquefies and shrinks away from the source of heat. The only way it can be made to burn is to have the heat source follow the retreating material. If the heat source is sufficient, and if it is maintained in contact with the EPS, it will when sufficient heat has been induced in it be set alight and will burn with "an intense black smokey flame". Once flaming is established in EPS, it is approximately twice as flammable as soft wood. Various chemical compounds, usually chlorine, phosphate, or antimony based, can be used as fire retardants in the manufacture of EPS for commercial purposes. When EPS burns, there is in effect a chain reaction, whereby there is first a breakdown of the polymer chains into individual molecules of styrene. These oxidise, giving off heat, which causes further decomposition and further burning. Chemical compounds used as fire retardants tend to slow down this decomposition and oxidisation process, and so retard the initial establishment of flame, but once a fire is established in EPS the retardants have little or no effect, and the material will under favourable conditions burn fiercely. If EPS in the form of Isolite is enclosed in a cardboard container, and the cardboard is set alight, the spread of flames in the cardboard is capable of acting as a constant heat source which may have the effect of setting fire to the Isolite.

Some discussion took place during the case about the meaning of the expression "self–extinguishing" in relation to Isolite. The expression is used in some of the promotional material which had been put out by OGP. For example, in a pamphlet which is Attachment No 3 to Mr Bowler's report (BP34), the brochure being dated April, 1970, this is said, "only types of styrene foam which fulfil the self–extinguishing requirements of Appendix F of A S K 156–1965 are included". Dr Cherry's evidence was that the expression "self–extinguishing" would have no ordinary meaning apart from the context of a standard, by which I take him to have been referring to the abovementioned standard. He said that the expression implies that when a source of heat is removed the flame will go out, but in the absence of a standard, that is a meaningless concept because whether the flame will go out or not for a given material depends upon the size of the flame when the source of heat is removed. He agreed that to say this substance is self extinguishing would give little true indication of its combustible qualities, in a real life situation such as a fire. I accept Dr Cherry's testimony on this matter. However, there is no evidence that any of the parties was misled by use of the expression "self–extinguishing" in any of the promotional material put out by OGP.

The following description of the combustion characteristics of foamed (ie expanded) polystyrene, taken from a pamphlet which was produced by the defendant OGP at the request of the plaintiff BPA (in Action No 2860A of 1982), and tendered in evidence by the latter (Exhibit BP31), contains material which is accurate, according to the tenor of the evidence given by Dr Cherry, and is useful to add for the present purpose to the layman's description set out above. The pamphlet is dated July 1974, is headed "Loss Prevention Data, Polystyrene for Construction", and was attached to a memorandum sent by the then marketing planning manager of OGP, Mr CH Buchanan, to another executive within his company. Mr Buchanan's attached memorandum is dated 18 November 1974. The material is as follows:–
"Since the forming of the foamed polystyrene is a mechanical process, there is no chemical reaction which might result in spontaneous ignition. While the flash ignition temperature of polystyrene is 600^o to 650^o F (315^o to 345^o C) and the self–ignition temperature is between 900^o and 950^o F (485^o to 515^o C), this can be misleading. The product will decompose at lower temperatures forming a flammable liquid. This occurs over a range of temperatures usually below 300^o F (150^o C).
The product can be ignited by an open flame and will burn in the presence of a flame generated by other fuels. The material tends to shrink away from heat sources prior to ignition, so the heat source must either be sufficiently large or must follow the shrinking material to ignite it. For this reason, attempts to ignite the polystyrene with a stationary match or Bunsen burner may not be successful. However, in a larger fire, such as one involving building contents, the heat source will be sufficiently large to sustain burning of the polystyrene foam.
Flame retardant additives do not usually have significant effect upon the burning characteristics except that they may slightly delay ignition. In certain applications, such as approved ceiling panels (discussed below) these additives have a beneficial effect.
Polystyrene foam forms a flammable liquid as it burns. When its surface is left exposed, flame may spread rapidly across it. For this reason automatic sprinklers are not always effective in confining the fire to a small area.
The material does not tend to smoulder, so slow burning internal fires which could later kindle another fire outbreak are not characteristic of foamed polystyrene.
In addition to its severe burning characteristics, the material also gives off a very dense black smoke containing oily, sooty particulate matter in suspension."

(d) Occurrence of the fire

I propose to set out the evidence of the two relevant witnesses in some detail.

Mr Lillico said that from the position where he was working, earlier described, approximately under the centre line of the building, about 15 metres from the stack of cartons, he looked towards them and "saw a glow and a lick of flames around the north eastern corner" (of the stack). He called out there was a fire, and ran towards it. From the position he was in when he saw this, he would have been at an angle to the stack, so that he could see the whole of its northern face, but none of the eastern face. As far as he knew, he was the first to see the fire and give the alarm. At the Inquest, he said that he had first seen "a fairly small glow compared to the size of the room ..... down the eastern end especially on the wall that covers the gable end above the ceiling". He saw this glow momentarily, before he saw the first flame. He saw the glow about two or three minutes after returning to the ceiling from ground level. The next thing he saw after the glow was a small flame, which seemed to be around the corner of the stack of cartons.

In evidence at the trial, Mr Lillico said that by the time he got to the fire, it had spread across the top of the carton stack. Another man, Jack Aherne, was already there using a fire extinguisher on it, but with no effect. It appears Lillico was wrong about that, because the preponderance of evidence indicates there was only one extinguisher in the roof void, and I accept Mr Anderson's evidence that he was using that. However, Anderson was further from the fire than Lillico when Lillico first saw it, and it is hardly likely that Anderson arrived at the fire first and was using the extinguisher before Lillico got there. Lillico also said in evidence that he thought there were two fire extinguishers in the roof void. This is one of the aspects of evidence which might have been elucidated if the other men in the roof void at the time had been called.

Mr Lillico said at the trial, "It was a ferocious fire and there must have been a draught from the eastern end which was more or less open, a draught coming through and it just swept across the top of the cartons and it would have been quite a big flame by then, two or three feet high and the full length of the stack". This was in answer to a question asking him to describe the fire when he arrived at it. On the western and southern ends one or two men threw half a dozen or so cartons off the stack, but the heat drove them back. The cartons had become a flaming mass, but he could not remember them collapsing. He realised it was impossible to extinguish the fire and soon realised also that the ceiling itself was on fire because of some small flames coming up through joints in the ceiling. They all realised they had to get out, and descended to the ground down the western side.

Cross–examined further at the trial about these flames coming through the ceiling, Lillico said that when he saw them he decided it was time to leave, because it looked as though the ceiling was on fire. The flames were coming through butt joints between panels of Isolite which constituted the ceiling (the panels being faced with thin sheet metal on either side), where the butt joints had not been capped. The flames were coming through at a point north of the stack of cartons, about six or eight feet to the left of where he was standing; but he did not see any flames or fire joining the stack of cartons and those butt joints on top of the ceiling.

At the Inquest, Mr Lillico had given an account of the fire which was generally similar but more detailed than that which he gave at the trial. At the Inquest, he said that after seeing the glow, he saw a small flame which seemed to be around the corner of the stack of cartons (ie on the eastern end). He thought the flame might have been between the cartons and the edge of the ceiling, but it got bigger and he realised that the stack of cartons was on fire. The area of flames grew quickly and in a very short time the whole stack of cartons was involved. He "headed towards the stack fairly quickly as I could", and "by the time I got over to it the flames had spread right along the top of the stack and someone else was there with a fire extinguisher". Others were trying to pull boxes off the stack, but he, held back by the flames, was standing there, and then saw flames coming through the butt joint which did not have a cap on it. This was on the northern side of the stack of cartons. Flames were leaping through this (area between the butt joints) for quite a distance. They all decided to go, and when they left the flames were still coming through the open butt joints. In cross–examination, he said that when he saw the first flame it was towards the bottom of the stack of boxes. It had only taken him a few seconds to get from where he first saw the fire to the stack, and by the time he arrived the fire was burning fiercely along the top and along the far (eastern) side. The total time from when he arrived at the stack and when they all left was between two and five minutes. This witness marked various relevant points on a plan (P17) at the Inquest, and although these markings in my view could not be relied upon for accuracy, they do generally accord with his oral evidence. He also said that there was dense black smoke in the roof void before he left, and that the closest distance between the flames coming through between the panel joints and the stack of boxes was between 10 and 12 feet (to the north).

Mr Anderson's account of the fire was in substance as follows. (His deposition made at the Inquest was not tendered). From his position approximately under the centre line of the building above Stage 2, about 15 metres from the stack of cartons, he heard a shout of "fire" and looked towards them. He saw that the northern side of the boxes was alight. The flames seemed to be coming from the base of the boxes, and they travelled very quickly, spreading upwards. There was a fire extinguisher nearby, and as soon as he saw the flames he grabbed it and headed towards the boxes. He applied the fire extinguisher to the flames, exhausting its contents, but without any noticeable effect. One or two others were trying to pull boxes away from the stack. However, the fire was burning strongly and spreading very rapidly. He was able to get very close to the stack but was driven back by the intense heat, and then he remembers that the roof began to explode. At one point he was within five or six feet of the boxes but was driven back to about 15 feet or so from the stack. When the roof started to explode, the men evacuated. He shut off the valves to the area to prevent any chance of explosion from the ammonia pipes.

In cross–examination, Mr Anderson amplified his description of the fire. By the time he arrived at the stack it had spread over most of the surface of the boxes. He agreed that he had said at the Inquest that he had first seen the fire at the north east corner of the stack. Describing the fire when he arrived at it, or within a very short time afterwards (meaning, I infer from the general tenor of his evidence, within a minute or so), he said, "It's like a furnace that sort of fire". In further cross–examination he said that when he arrived at the fire:–

".... the whole face of the boxes were alight, including around to the western side of the boxes were alight, as well.
Even in the short time it took you to get from picking up the extinguisher to getting over there? ......Yes, oh yes, the breeze just fans them very quickly, like a grass fire I suppose would be the best way to describe it."

Asked what he meant by the breeze, he said:–
"Oh well there's a natural draught in those styles of buildings ...... You don't feel it when you are working there. I mean, you can sort of – it's sort of nothing you really notice until something like this occurs and then you can – it seems to sort of create a natural draught, I would say, the fire. ....... They are an open type building, even though there are gabled ends, you do tend to get a bit of wind movement up from underneath the eaves, because they are open."

Asked whether he was assuming the existence of a draught or whether he remembered one, he said in effect that, remembering seeing the way the flames were travelling, he imagined they would have to be forced by a draught of air. This witness also said that as far as he could tell it was only the edges of the boxes that were burning; but as I shall say later, I think that impression he had is probably wrong.

As to the period immediately before and after the fire, Mr Lillico said that he had gone to ground level shortly before the fire to check why there was no power in his welding machine, and while on the ground he looked in room 5 and saw no–one there. There was no sign of fire and all appeared normal. He had only been back on the ceiling for five to ten minutes when the fire broke out. After leaving the roof void, he walked to the north side of the building and there saw that the fire had broken through the roof and a great deal of smoke was being generated. The centre of the fire seemed to be at the eastern end of the Stage 2 extension.

(e) Burden of proof

The principal issue in the case is whether it should be inferred on the evidence, on a balance of probabilities, that the fire was caused by products of a welding operation negligently carried on by an employee or employees of WS in the roof void. Although there are associated allegations against various of the parties, this remains the basal issue.

The onus of proving the affirmative of that issue remains throughout upon the plaintiff. In considering the question of fact, whether it is more probable than not that the fire was caused by a product of welding, the court should take into account all relevant circumstances which tend for and against that proposition. In order for a party to establish the affirmative of the proposition, it is not necessary that any particular fact which would tend to prove it must itself be proved on a balance of probabilities. For example, it is not necessary for the plaintiff BPA to prove against the defendant WS, as a specific component of discharging its burden of proof, that the fire started in a part of the stack of cartons of Isolite which was within range of welding particles. In order to discharge the onus, the relevant circumstances considered overall must be such as to warrant a reasonable mind reaching an affirmative conclusion that the fire was caused by a particle or particles emitted by the welding – compare Tubemakers of Australia Ltd v Fernandez (1976) 50 ALJR 720; Dahl and Another v Grice [1981] VR 513, and see also Rhesa Shipping SA v Edmunds [1985] 1 WLR 948. The factual context of the first two of those cases is rather more specialised than in the present instance, in that they deal with the question whether, in relation to a workers' compensation claim, the absence of expert medical evidence tending to prove a probability rather than a possibility that a particular injury or disease is causally connected with a given event, an affirmative conclusion on that question may nevertheless be reached by a tribunal of fact having regard to overall relevant circumstances; but nevertheless they do illustrate the general proposition. In relation to causation and the problem of judging probabilities, the present case is more like Rhesa Shipping SA v Edmunds (supra).

Dixon CJ in Jones v Dunkel (1958–59) 101 CLR 298, expressed in the following way the nature of the court's task in deciding whether a proposition has been proved on a balance of probabilities in an action based on negligence:–

".... we are not concerned with a choice among rival conjectures. In an action of negligence for death or personal injuries the plaintiff must fail unless he offers evidence supporting some positive inference implying negligence and it must be an inference which arises as an affirmative conclusion from the circumstances proved in evidence and one which they establish to the reasonable satisfaction of a judicial mind. It is true that 'you need only circumstances raising a more probable inference in favour of what is alleged'. But 'they must do more than give rise to conflicting inferences of equal degree of probability so that the choice between them is mere matter of conjecture'. These phrases are taken from an unreported judgment of this Court in Bradshaw v McEwans Pty Ltd (unreported, delivered 27th April 1951) which is referred to in Holloway v McFeeters (1956) 94 CLR 470, by Williams, Webb and Taylor JJ The passage continues: 'All that is necessary is that according to the course of common experience the more probable inference from the circumstances that sufficiently appear by evidence or admission, left unexplained, should be that the injury arose from the defendant's negligence. By more probable is meant no more than that upon a balance of probabilities such an inference might reasonably be considered to have some greater degree of likelihood' (ibid, at pp 480, 481). But the law which this passage attempts to explain does not authorise a court to choose between guesses, where the possibilities are not unlimited, on the ground that one guess seems more likely than another or the others. The facts proved must form a reasonable basis for a definite conclusion affirmatively drawn of the truth of which the tribunal of fact may reasonably be satisfied."

In the present case it is clear that the distance over which hot particles emitted by an arc welding process may travel, the distance from the "weld pool" to any point in or on the stack of cartons where the fire might reasonably be thought to have started, and the elapsed time between the end of welding and the beginning of the fire, are important elements to be considered in relation to whether the burden of proof has been discharged. But they are not the only elements. The facts that the cartons of Isolite were stacked and situated where they were, that a welding operation did take place (assuming that be accepted as a fact) within a few metres of the stack, at a relatively short time before the fire was seen, that the fire itself occurred, that no other plausible explanation for the fire has been put forward (if that be accepted as a fact), that there is evidence, and it is well known, that welding operations at industrial sites, where combustible material tends to be present, are likely to cause fire unless precautions are taken, are all relevant facts.

The defendant WS has in this case called evidence with the object of inducing a persuasion that the time and distance elements were such that it was impossible, or so improbable as to verge upon impossibility, that the fire could have been caused by a welding product. Further, it claims to have succeeded in showing that; and so it is necessary to examine closely the evidence regarding those two elements.

(f) The time element

Evidence as to the performance of welding by Mr Ian Whyman, an employee of WS, and the time when that was done in relation to the beginning of the fire, is contained mainly in certain interrogatories tendered. Exhibit BP11 is an affidavit by Mr JR Beard, Managing Director of WS, on whose behalf it is made. It contains answer No 18, as follows:–

"Some time before the fire, Mr Ian Whyman was welding a plate by standing on the Burnie Freeway end of the RSJ closer to room 4, being one of the pair of RSJ's in room 5 in position to support an FDC unit at the Bass Strait end of room 5."

There is also No 20, the substance of which is to say that when Mr Whyman last operated the welding machine before the fire, the approximate distance between the weld point and the nearest part of the cartons of Isolite was five to six metres. No 23 says that Mr Whyman was performing a fillet weld on a join between a plate and a vertical U–beam at eye height from where he was standing on the (relevant) RSJ.

Exhibit GJ3 is also an affidavit by Mr Beard, made on behalf of WS, exhibiting answers to certain interrogatories. The effect of question 10(c) and (d) is to ask when the welding operation carried out by Mr Whyman commenced, and when it ceased. The answer is as follows:–

"W & S is unable to be any more precise than to refer to the sequence of events recounted by the witness Mr Ian Whyman at DP 147 and DP 156 and by the witness Mr Kerry Aherne at DP 250."

These numbers refer to pages of the depositions taken at the Coronial Inquest. In these pages, Mr Whyman says that after he finished welding the relevant plate, he was asked to go up to the western end of the building to give some advice to two of the other workmen. He thought that he had not actually finished the weld when they called and asked him to come and help them. He was in their company when he saw the fire. He was asked "How long after you first left the fillet weld to go down to Ricky and Steven was it that you noticed the first sign of fire?" He answered "I was down there about 10 or 15 minutes before somebody yelled out". At p250, Mr Aherne was asked, "Approximately how long was it between the time you last saw Mr Whyman doing any welding in that location you have just mentioned and the time that you learned that there was a fire?" The answer is, "Between 15 and 20 minutes".

There is also evidence from Dr Barrett, one of the experts, as to the time it would probably take for a smouldering piece of weld particle to ignite a carton of Isolite. Dr Barrett is, in retirement, a fire investigator. He is highly qualified as a fuel technologist and chemical engineer. The witness said that most of his working life has been associated with burning of many kinds of substances. His expertise includes the manner in which and the temperature at which ignition can take place in various kinds of fuels. He said that particles of the kind emitted by the arc welding process (of which I shall say more later) were capable of igniting boxes of expanded polystyrene, but it would be "most improbable" that any one particular particle would do so. For that to happen, it would probably be necessary for the particle to lodge in a situation where conditions for ignition were very favourable. The ignition temperature of cardboard would probably be lower than that of Isolite. A particle emitted from a welding process would probably require to be "quite a bit hotter" than 400^o C to cause ignition of cardboard, because it would have to transfer some of its heat energy to that material. He thought that probably 300^o to 360^o C or thereabouts would be necessary to cause ignition of Isolite. Dr Barrett said, "You need to get ...... a physical thing that surrounds the particle, the hot particle, so that the heat losses from the particle go into the surrounding material and don't escape, and under these conditions, your expectation of an ignition would be quite a deal greater." He was asked, "Now you heard some discussion yesterday about the delay that may occur after the particle lodges in the crevice. What do you say about that? Can there be a delay?" The answer was, "Yes, I wouldn't have expected a very long delay. I think the system would be such that it would be either 'go' or 'no go' within a couple of minutes; something like that."

I accept these opinions as being probably correct, and as meaning that if a particle of molten metal, eg spatter, with sufficient temperature and heat content, lodged in a crevice in a cardboard carton or between cartons in a situation where necessary conditions for ignition were favourable, and ignition took place, it would most likely take place within a matter of minutes, say from two to five or six minutes or thereabouts, though it is not possible on the evidence to be precise. I infer that it would take some further time, once ignition had started in a pile of cardboard cartons containing Isolite, for the fire to build up so as to be furnace–like.

(g) The distance factor

Whether any part of the stack of cartons was close enough to the weld point for a particle or particles from the weld pool to have started the fire, or for a hot electrode stubb flicked by the welder to have started it, depends, on the evidence, mainly upon the size and conformation of the stack, where in the stack the fire commenced, and how far such objects may be projected.

It is not possible to infer with precision how far the eastern end of the stack was from the eastern edge of the ceiling. Mr Lillico's is in substance the only evidence about it. He said on one occasion, about ten feet. Once or twice at the Coronial Inquest he estimated the distance as six to eight feet or thereabouts. On his evidence, the distance would be about ten feet or less; but I have already indicated I do not place much reliance upon the accuracy of his estimates of distance and the like. How far the stack extended back towards the western side depends upon the number of boxes in the pile and how they were stacked. The number could have been anywhere between 20 and 30 or thereabouts. The burden of the evidence of Messrs Lillico and Anderson is that the stack was reasonably neat and rectangular, but nevertheless the cartons were not closely aligned. They were roughly stacked, and I infer that there were probably numerous ledges, crevices, and angles between the alignments. That is, it is probable that there were many places where a molten particle might lodge and find favourable conditions for ignition.

If the fire started at the eastern end of the stack of cartons, which is the proposition argued by WS, then, since the position of the weld point is known, and if the eastern end of the cartons can be inferred to have been about ten feet or less from the edge of the ceiling, the distance which a particle would have to fly or be projected in order to start the fire would be some seven metres or a little more. It is certainly, on any of the evidence presented, very unlikely that any piece of spatter or slag could travel so far from the weld point. That a spent electrode could be flicked so far by the welder when disposing of it from his hand piece, as the evidence is some welders do, is also improbable, but the possibility cannot be eliminated. Whether such an electrode could be hot enough to cause ignition is unexplored by the evidence, but it was said they are very hot when they leave the hand piece.

(h) The content of expert evidence

Various expert witnesses gave evidence about the kinds of particles emitted by the arc welding process, and the distances they may be projected. Three kinds of material are emitted. Sparks are tiny pieces of molten metal, having a mass of about 1 milligram when solidified. Ordinarily they are solidified by the time they land on a surface, and can be collected as small spherical particles of metal. Spatter consists of larger pieces of molten metal, with a mass of about 100 milligrams or thereabouts, which either break up into sparks, or flatten out when they strike the ground.

The third product, slag, is a crust which forms above the weld, and consists of oxidised metal and flux residues. It is inorganic; not metal but a ceramic. It is chipped away with a hammer by the welder when the weld is done, or sometimes while the weld is progressing. How far a piece of slag will fly depends upon the manner in which the welder strikes it off; the force used, the angle of the blow, and the like. Slag can be very hot when it comes to rest on a surface.

The "Bowler Report", exhibit BP34, was treated in the case by some of the other experts as a kind of basic document and starting point in relation to welding. Mr Bowler, who is the Hydro Electric Commission's metals technology engineer, and a special adviser to the Government of Tasmania on technology, engineering and development, was commissioned by the Coroner to investigate and report on the fire in question. His report was tendered in evidence, although not all parts of it are to be taken as admissible. I do not stay to examine which parts are admissible and which not, a matter which was not in contention. In substance, the whole of the report is admissible, other than any opinions expressed as to the cause of the fire, or opinions as to the parts played by any of the men working on the site at the time, or any factual material concerning the fire. Any statements within those categories are treated as assumptions only. Mr Bowler's basic qualifications are local, obtained at the Hobart Technical College, but he has had interstate and some international experience on various committees and bodies connected with welding and research into welding. He has had extensive experience within the Hydro Electric Commission with welding and as a supervisor of trainees in welding, and the like. His report in substance was accepted by the other expert witnesses, and was not attacked for any alleged deficiencies in relation to its technical content.

Mr Bowler carried out certain fire tests on cardboard cartons containing Isolite insulation. He said the tests showed that such a carton was easily ignited by a small ignition source such as hot metal and slag from a welding operation. He said the burning cardboard box acts effectively as a large heat source, thereby causing pyrolysis of the EPS insulation and ignition of the gaseous pyrolysis products. His report gives a very short description of a test (No 15) which he said showed this result, but it involved oxy–acetylene cutting. In his experiments, a carton was "soon reduced to a fiercely burning residue of liquid", giving off thick black smoke. He expressed the opinion (p16 of the report), which I take to be as a result of his tests and his experience, that high temperatures resulting from burning cartons of Isolite standing upon Isolite panels of the kind used in the ceiling of the cool store would be capable of melting the Isolite panels, and causing (in effect) the ceiling to collapse. He also said that the tests indicated that the aluminium surface of an Isolite panel could be perforated in that way, allowing direct access of flame into the core of the panel. Mr Bowler also said in the report (p17) that "electric welding creates sparks and particles of metal and slag at high temperature". These particles can travel quite long distances while retaining sufficient heat to cause combustion. The maximum scatter distance of the particles is a function of the welding procedure and the height of welding operations above floor level. Simulated tests had shown that in certain circumstances an arc welding operation could result in an occasional particle travelling six metres or so. However, most particles would be contained much closer to the welding point.

Mr Bowler also obtained samples of the grades of polystyrene panels and insulation understood to have been used in the cold store, tested them, and found that they conformed to the relevant Australian Standards for self– extinguishing polystyrene. He said in oral evidence that he found it was very difficult to create and sustain a fire in Isolite, though in his experiments (as illustrated in the report), he succeeded in doing so. However, the sources of ignition which he used did not except in the case of Test 15 consist of products of a welding process, and did not correspond with them. I do not take his experimental results as providing useful evidence as to whether and when products of the welding process are capable of igniting cardboard cartons of Isolite.

In oral evidence, Mr Bowler expressed the opinion that burning cartons of Isolite situated on top of the ceiling constructed of Isolite panels would in fact melt the Isolite in the sandwich construction of the roof panel, thus reducing the capability of the roof panels to sustain (weight). In addition it would be likely that the molten Isolite would run out between the joints, and certainly the metal facing would peel away from the surfaces of the modular panels. (As a result) the ceiling would eventually collapse. Mr Bowler also said, and I accept, that in his view the distortion he observed in portal 12 and the southernmost RSJ after the fire, the damage being more severe in that area than in other areas, was caused by "the intense fire from the stack of boxes before the roof caved in". By "roof" he meant the ceiling panels of the cold store. I treat as an assumption his statement that the ceiling caved in.

Dr Cherry treated Mr Bowler's report as a starting point, and conducted experiments with the object of seeing whether products of a welding operation such as he assumed to have been carried out by Mr Whyman could have caused a fire similar to the one in issue. Stated more accurately, the object was to try to ascertain whether a globule of molten metal or a piece of slag produced by the welding process would have sufficient thermal energy to ignite the system consisting of lengths of Isolite used for insulation, enclosed in cardboard cartons of the kind used in this case. The experiments are described in detail in the evidence, and shown in full in a video tape which was tendered and shown during the trial. The experiments were a series in which a welder performed welding operations of a kind and in a similar situation to that assumed to have been done by Mr Whyman, at different distances from stacked cartons of Isolite. The products of the welding process, how far they travelled, whether they caused charring or fire on flat surfaces of cardboard or in cardboard cartons, all were carefully noted, collated, and where possible collected. However, one aspect of his experiments which I think reduced the value of Dr Cherry's results in respect of distances to which welding products can be projected was that the welder stood on what amounted to a metal platform, upon which much of the welding product fell, with the result that it tended to be projected outwards in an explosive kind of way (as seen on the video). This may have extended or altered the distances to which particles might fly from the weld pool to ground level. Subject to that comment, I shall summarise some of the more closely relevant aspects of the results obtained by Dr Cherry.

Dr Cherry found particles of slag up to 4.5 metres away from the weld point. He found that 32% of the slag came to rest within a metre of the welding area; 30% between 1 and 2 metres, 28% between 2 and 3 metres, and 10% between 3 and 4 metres. His conclusion about slag was that pieces can travel for distances up to and beyond four metres, and that 20% of all the slag in his experiments travelled more than 3 metres from the welding point. He also concluded that if a piece of slag travelled to a location four or more metres away, and fell into a crevice or the like between cardboard boxes, it could start a fire.

Although charring occurred on a number of the cardboard surfaces upon which particles fell, Dr Cherry in 20 or so experiments was not able to start a fire, except in two cases. The two exceptions were when he had, first, some empty cartons stacked virtually under the welding point, and secondly, some cartons filled with Isolite stacked in about the same position, and on each occasion he and another person scraped pieces of slag and spatter which fell onto the horizontal surface of the carton into a crevice between cartons. Dr Cherry said in oral evidence that within two or three minutes the whole structure of cartons containing Isolite was well alight.

I make the following comments about the "fire results" obtained by Dr Cherry. I accept his experimental results as showing truly what happened. But, exercising what I hope is a broad commonsense judgment, I do not accept these results as being by any means necessarily exhaustive of what could happen if and when some of the molten products emitted by the arc welding process land in and among cardboard cartons of Isolite stacked as the cartons in question were before the fire. In the nature of things, short term experimentation could not exhaust the range of reasonable possibilities in the real situation, having regard particularly to the practical impossibility of duplicating exactly all the conditions which might apply in the latter. For example, the kinds of surface on which the particles were collected in the experiments were not the same as the surface where the real fire occurred. There, particles might have bounced off the top of the RSJ which I have inferred was close to the cartons. Some particles might have been projected over exceptionally long distances. A particle lodged in a crevice might have been fanned by a draught or current of air. Even if it be assumed to be extremely improbable that any given particle or spent electrode stub would be projected far enough, carrying sufficient thermal energy, and would lodge in a situation where conditions would be sufficiently favourable for ignition (though I do not make such an assumption or finding), that would not necessarily preclude an inference on the balance of probabilities that products of welding caused the fire. Such a conclusion would have to be formed or not upon a consideration of the whole range of relevant circumstances.

Dr TR Blackburn is a senior lecturer in the Department of Electrical Power Engineering at the University of New South Wales. He has a Doctorate of Philosophy in Physical Sciences from Flinders University, and his specialty is discharges caused by the flow of electric current through gases. Dr Blackburn after obtaining his doctorate spent five years studying the fundamental properties of electric arcs, including welding arcs, arcs in furnaces and arcs in electrical switchgear. He has done some welding, but is not an expert welder himself. He prepared for making a report and giving evidence in this case by studying various technical papers to which he was referred by the Australian Welding Research Association and some of its officers. Two papers to which he paid particular attention and was prepared to rely upon were, one by Japanese scientists Higawa and Ikeida, and the other by Swiss authors Bura and Runchweiler (the spelling here may not be correct). The witness' written report was not tendered in evidence. He said that in preparing the report he was concerned principally with determining whether the temperature of particles emitted during the welding process could be sufficient to cause ignition of flammable materials after travelling a distance between 4 and 6 metres. He was given certain information and asked to assume it to be correct, and he made other assumptions of fact based upon information provided in the two technical papers mentioned above, which based upon his experience he was prepared to accept as correct.

Dr Blackburn said that the information he needed in order to perform his calculations was, first, the minimum temperature required to ignite flammable cellulosic material (in this case cardboard). He took that to be up to 400^o C, which temperature he obtained from a technical note published by the Australian Welding Research Association. (Dr Cherry referred with approval to that technical note). He also needed to know the temperature of the weld pool, in order to know the temperature which particles would be at when they were emitted. From two technical sources, which he named, he took that temperature to be 2,200^oC, in normal steel welding. He had a certain amount of experience of his own which tended to support those assumed values.

Relying upon the Japanese paper, the witness assumed that particles leaving the weld pool could travel "at least up to 7 or 8 metres from the heights – welding heights that were involved here and up to 20 metres from much greater heights and so that paper showed that these distances could be traversed ...". It came out later, however, that the minimum height of the weld above ground being considered by the Japanese paper was 8 metres, and so Dr Blackburn's evidence was not useful on the question of how far particles emitted from the weld pool in Mr Whyman's welding process might travel; but that was not the purpose of his calculations and evidence, as stated by him. A further assumed fact which he took into account was that molten particles may leave the weld pool at an initial velocity of up to 9 metres per second. He said this was a measured figure which he had obtained during experiments which he had himself conducted or taken part in. Dr Blackburn stated and expounded a number of considerations which he had taken into account in arriving at his results, and in the end stated the following conclusion:

"I calculated the – given the velocities that we had measured under distances involved, 6 to 7 metres, I calculated that the time required for the particles to travel that distance would be only of the order of a couple of seconds, and in that time and given the life times of particles that I've calculated I concluded that after travelling those distances the temperatures of most of the particles would still be well above the 400^oC ."

Dr Blackburn was criticised by counsel for not having taken air drag or heat content (ie, thermal energy) into account in reaching his conclusions, but a reading of the transcript shows that he did not overlook those matters. However, my impression of the core of Dr Blackburn's evidence was that, having regard to the number of technical factors which he had to make a judgment about and take into account in his calculations (which themselves were not exhibited in evidence), even as a layman I was not persuaded that it was more probable than not that his conclusion was correct. A reading of the transcript between pages 1278 and 1295 will illustrate what I mean. I make that comment with all due respect to Dr Blackburn, because I have no reason to think from seeing and hearing him that he is other than a scientist of the highest repute and competence. The only positive finding I can make from Dr Blackburn's evidence, given that his evidence was not subjected to critical examination by any other expert witness in the case, is that it is arguable from a technical point of view that products of the sort of welding process conducted by Mr Whyman could be capable of retaining a sufficiently high temperature to ignite cardboard of the kind used here at a distance of 6 to 7 metres.

Dr Barrett, mentioned earlier, described technical aspects involved in the burning of fuels, including EPS. It was argued by Mr Barnard for BPA that evidence given by Dr Cherry and Dr Barrett disposed of any suggestion that the fire may have been caused by products emitted from welding carried out by a Mr Goninon, a welder employed by BPA, at ground level on the day of the fire. I accept that argument. No counsel seriously argued otherwise.

(Before leaving Dr Barrett's evidence, I should comment that on more than one occasion he said words to the effect that Isolite is very flammable, but it is clear from the context that he was using "flammable" consistently as meaning that Isolite burns strongly once it is set well alight. He did not mean that it is easy to set it alight in the first place, which is the dictionary meaning of flammable – see OED).

The next technical witness I refer to is Mr JC Arthur. He is a consulting mechanical engineer with about 40 years' experience in engineering, including some 4 years as a consultant. He has a Bachelor of Engineering degree from the University of Tasmania, a Diploma in Mechanical Engineering, and also in earlier years completed an apprenticeship as a fitter and turner. During much of his engineering career he has been involved with training welding apprentices, developing welding procedures, and with (I infer, the direction and supervision of) welding jobs involving high pressure vessels, high pressure pipe work, and the like. He has therefore had a great deal of experience with welding both theoretical and practical, and was well qualified to speak about safety practices and the like.

(The only thing which might seem worth noting here in relation to that passage is that it appears to be necessary to add in the effect of Turberville v Stamp (supra) in order to make clear the negligence aspect of the ancient rule). Edmund Davies LJ agreed with the judgment of the Master of the Rolls, and said that the Australian decision in McInnes v Wardle (1932) 45 CLR 548, and Balfour v Barty–King (supra), are illustrations of the relevant law. Phillimore LJ agreed, citing Black v Christchurch Finance Co Ltd [1894] AC 48 as a further example.

Emanuel's case (supra) was one in which the London County Council arranged for two houses it owned to be demolished and removed by an independent contractor. The contractor started a fire on the site to burn rubbish, even though the agreement provided that this should not be done. He was negligent, and sparks flew to the plaintiff's neighbouring premises, causing damage. The County Council was held liable. In Balfour v Barty–King (supra) an independent contractor negligently and unintentionally started a fire, by using a blowlamp to heat up some frozen pipes in order to free them. The occupier was again held liable. In both these cases the fire was not started intentionally, but resulted from what was referred to in Balfour's case as "a collateral act of negligence" – per Lord Goddard, at p504 (delivering the judgment of the Court of Appeal). The distinction is worth noting, because cases in which the person who lights the fire does so intentionally, such as McInnes v Wardle (supra) and Black v Christchurch Finance Co (supra), tend to be considered within the category of nuisance or under Rylands v Fletcher (supra), rather than under the old common law rule.

It may be observed also that in Sturge v Hackett (supra), Collingwood v Home and Colonial Store, Ltd [1936] 3 All ER 200, and Musgrove v Pandelis [1919] 2 KB 43, the law as to occupier's liability for the escape of fire, independent of nuisance or the rule in Rylands v Fletcher (supra), was stated in the same way as in Balfour's and Emanuel's cases (supra), but the former were not instances of the occupier being made liable as the result of negligence on the part of an independent contractor.

However, the law being, as it seems to me, thoroughly well established over a great length of time in England that an occupier is liable for the negligence of his independent contractor in that respect, as is confirmed by the well known text books (eg Salmond on Torts, 17th ed p333; Winfield and Jolowicz on Torts, 10th ed, p383; Clerk and Lindsell on Torts, 15th ed, p1222, paras. 24–33; Street, the Law of Torts, 6th ed, p260), the question is whether the relevant law is different in Australia. BPA, supported by WS, contends that it is different, citing the Victorian case, Pett v Sims Paving and Road Construction Co Pty Ltd [1928] VLR 247, which in turn relies upon the authority of an earlier High Court case, Bugge v Brown (1919) 26 CLR 110. These cases, however, in my view do not support the proposition for which they are cited.

In Pett's case (supra) the facts were that the defendant was an independent contractor making and paving a driveway through the land of a person (ie the occupier) who was not a party. The contractor brought on to the property a steamroller and bitumen melter, and sparks therefrom caused a fire in neighbouring premises, but without negligence. Irvine CJ said in his judgment that on those jury findings counsel for the plaintiff had moved for judgment, relying upon "the old doctrine of the common law as stated in Rolle's Abridgement, 'Action sur Case,' B1,3, and restated in some of the modern textbooks, whereby an absolute duty was thrown upon any person on whose land a fire was lighted, by himself, his servant or licensee to prevent such a fire escaping, inasmuch as the making of a fire involves the bringing on land of something not naturally there, and therefore the owner of the fire is bound to keep it in at his peril". His Honour then, after saying that the proposition so stated had been questioned by some modern writers including Sir John Salmond, said he need not refer further to those earlier cases, because of a statement by Isaacs J to which he referred in the case of Bugge v Brown (1919) 26 CLR 110. Irvine CJ then said:

"I have been unable to find any authority in modern times for the proposition of law that – except in two classes of cases – anyone who uses fire in any place and in any circumstances becomes an insurer of all persons who may be affected by its escape. The two classes referred to are where the doctrine of Rylands v Fletcher applies, and where the lighting or maintenance of the fire is shown, without the aid of that doctrine, to be a nuisance at common law."

His Honour then dealt with the case before him under each of those two categories, and held the plaintiff not liable. Thus, the way in which Irvine CJ stated and dismissed the absolute proposition in argument put to him as representing the common law rule does not affect or deny the law as stated in, for example, Emanuel's case.

I turn then to what Isaacs J said in Bugge v Brown (supra). There, the defendant owned certain grazing land. His employee lit a fire to cook his meal, against instructions as to where he was to light it. Fire escaped by his negligence to the plaintiff's land and caused damage. The court held that the employee was acting in the course of his employment, and the defendant owner was liable for the damage caused by his negligence. Isaacs J, in his judgment, said that counsel for the appellant had put forward three propositions for holding the defendant liable:

"First, he contended that the owner of land is liable for damage caused by any fire there in fact kindled or kept by his servant whether negligently or not, and whether or not in the course of his employment. This contention was rested on a case decided in 1401 – Beaulieu v Finglam."

His Honour then stated the other two arguments, and said:
"The first proposition the Court ruled against during the arguments. Whatever may have been anciently considered the true rule of the common law, the rigorous proposition so contended for cannot now be maintained."

Isaacs J then considered the case before him and resolved it on the basis of the third argument, which was that:
"An employer is liable where his servant lights or keeps a fire negligently but within the course of his employment."

His Honour's judgment was in substance wholly concerned with that third contention and so were the judgments of Higgins J and Gavan Duffy J Therefore, I do not regard either Pett's case (supra), or Bugge v Brown as denying the existence or validity of the rule of law as stated in Balfour v Barty–King (supra) and Emanuel's case (supra). Nor, when one considers other relevant judgments of the High Court of Australia, can it be found that the court has ever denied the rule of law as formulated in Balfour v Barty–King and Emanuel's case.

Thus, in McInnes v Wardle (supra) an independent contractor was engaged to fumigate rabbits on the defendant's rented property. The contractor lit a fire, which was a usual and ordinary method for fumigation and destruction of rabbits. The fire spread to a neighbour's land, causing damage, and it was held that the occupier, the employer of the independent contractor, was liable. The court followed and applied Black v Christchurch Finance Co (supra). Gavan Duffy CJ and Starke J said this in their joint judgments:

"The decision of the judicial committee in Black v Christchurch Finance Co settles, in our opinion, the principle that an occupier of land is liable for damage lighted in dangerous circumstances by an authorised person, whether servant or contractor, notwithstanding that the conditions of authority have not all been complied with or have been abused. 'The lighting of a fire on open bush land, where it may readily spread to adjoining property and cause serious damage, is an operation necessarily attended with great danger, and a proprietor who executes such an operation is bound to use all reasonable precautions to prevent the fire extending to his neighbour's property ... and if he authorises another to act for him, he is bound not only to stipulate that such precautions shall be taken, but also to see that these are observed, otherwise he will be responsible for the consequences' (Black v Christchurch Finance Co)."

Black's case (supra) was argued and decided in nuisance. The Judicial Committee, whilst their advice dealt mainly with the facts, cited Hughes v Percival 8 App Cas 443 for the proposition which the joint judgment of the Chief Justice and Starke J in McInnes v Wardle extracted and relied upon, as mentioned above.

Dixon J (as he then was) in McInnes v Wardle (supra) also dealt with the case on a nuisance basis. His Honour cited Rainham Chemical Works Ltd v Belvedere Fish Guano Co [1921] 2 AC 465 for this proposition:

"... the duty of an occupier to take care that his land is so used and the operations carried out upon it are so managed that his neighbours are not exposed to injury by exceptional dangers is not confined to dangers arising from acts of himself and his servants."

Everett J and McTiernan J in separate judgments followed and applied Black's case.

Several other cases decided by the High Court of Australia, dealing with the liability of an occupier of land for the escape of fire caused by the actions of his servants or agents, are relevant in this context but I need mention them only briefly. Hazelwood v Webber (1934) 52 CLR 268 was decided under the principle of Rylands v Fletcher (supra). The appellant had lit a fire on his land to burn off stubble, which was an ordinary farming operation; but a tree stump ignited and smouldered for five days, and then a high wind caused the fire to spread to a neighbour's land. The headnote rightly states that the court held that the burning of vegetation in the open in midsummer is not a natural or ordinary user of the land, and therefore the appellant was liable independently of negligence for the consequences of escape of the fire – see in particular the joint judgment of Gavan Duffy CJ and Rich, Dixon and McTiernan JJ, at pp277–279.

Wise Bros Pty Ltd v Commissioner for Railways (NSW) (1947) 75 CLR 59 was a case of a deliberate use of fire on the defendant's land, in that the defendant had a factory in which he manufactured sheep food, and in the course of doing that produced steam by using a portable boiler and fire box standing outside the factory. A fire broke out on his premises and spread, destroying not only his own premises but certain property of the Railway which was on a siding within the defendant's premises. The case was dealt with under Rylands v Fletcher (supra), following Webber's case (supra), the question being whether the use of fire in the circumstances was a natural or non– natural user. A new trial was ordered on the basis that certain evidence relevant to this point had been erroneously rejected.

An earlier relevant case was Whinfield v The Lands Purchase and Management Board of Victoria and the State Rivers and Water Supply Commission of Victoria (1914) 18 CLR 606. There, employees of the Lands Purchase Board camped on land owned by the Water Supply Commission, without having obtained permission to do so. One lit a fire for cooking purposes, and by reason of his negligence it spread to adjoining land and caused damage. No negligence was found. The case was discussed by members of the court in separate judgments under Rylands v Fletcher (supra) and nuisance generally. In substance it was held, affirming the judgment of the Full Court of the Victorian Supreme Court, that neither the Board nor the Commission were liable under either head.

Lastly, Goldman v Hargrave [1967] 1 AC 645 was decided entirely within the field of nuisance law. The facts were that a fire had been started high in a tree on a grazier's property accidentally, having been struck by lightening. The tree was felled, and the grazier (ie occupier) could have put the fire out within a day or so by taking reasonable steps, but deliberately did not do so. A high wind came up, causing the fire to spread to a neighbour's land and cause damage. The occupier was held liable in nuisance, he having a general duty of care in relation to hazards occurring on his land, to remove or reduce them in accordance with his knowledge of the hazards, his ability to foresee the consequences of not checking or removing it, and his ability to abate it. The Judicial Committee, incidentally, approved the application of Filliter's case (supra) and Musgrove's case (supra) as to the application of the 1774 Metropolitan Fires Prevention Act.

None of the above authorities as I understand them over–rules the law as stated (and for convenience referred to) in Emanuel's case (supra), and I conclude that this law is applicable in Tasmania and in the present actions. The rule is that an occupier of land is liable for damage caused by the spread of fire from his land caused by the negligence of his independent contractor. Since I have found that WS was negligent in this regard, and that it was an independent contractor employed on its premises by BPA, I must hold that BPA is liable to GJ under this head.

I turn to the argument that BPA is liable to GJ under the rule in Rylands v Fletcher (supra). I have no doubt it would be possible to establish liability under this head, separately from the common law rule above discussed, provided it was an appropriate case for application of the rule. This proposition is too well established to need exposition (see generally, Fleming on Torts, 6th ed, ch16). However, in my opinion liability cannot be established against BPA on this basis in the present case, because welding on the defendant's premises was not a non–natural user. Welding, after all, is a perfectly ordinary operation which is performed on countless occasions every day all over the world within buildings on industrial sites, and it can be performed with adequate safety and security provided proper precautions are taken. That is to say, it is an ordinary and usual way of using property. It is not a "special use, bringing with it increased danger to others" – Rickards v Lothian [1913] AC 263. The cases both in England and the United States, but particularly in the latter, abound with examples of industrial operations of a generally similar kind which have been held not to be non–natural user – see for example, Fleming, The Law of Torts, 6th edn pp308–310; Prosser and Keeton on Torts, op cit, pp545 ff.

(b) In Negligence

The case argued by GJ against BPA on the basis of negligence relies principally upon the decision of the High Court of Australia in Stevens v Brodribb Sawmilling Co Pty Ltd (1986) 160 CLR 16. That is to say, notwithstanding that, as is common ground, WS was an independent contractor engaged by BPA, nevertheless the nature of the work and the relationship of BPA to WS and GJ during the course of it was such as to create a duty of care on the part of BPA towards GJ not different from what it would have been if WS had been an employee of BPA. In Steven's case (supra) the court held that a sawmilling company which engaged a number of independent contractors to carry out various parts of its business operation of removing timber from a logging area to a mill, including felling trees, moving logs, constructing landings and ramps, and carting logs from the logging area to the mill, was as much under a duty towards two of those contractors to provide a safe system of work, in which they could act interdependently in carrying out their parts of the overall operation, as the company would have been if they were its employees. Mason J (as he was then), with whose judgment Brennan and Deane JJ agreed on this aspect, expressed the principle in this way:–

"Although the obligation to provide a safe system of work has been regarded as one attaching to an employer, there is no reason why it should be so confined. If an entrepreneur engages independent contractors to do work which might as readily be done by employees in circumstances where there is a risk to them of injury arising from the nature of the work and where there is a need for him to give directions as to when and where the work is to be done and to co–ordinate the various activities, he has an obligation to prescribe a safe system of work. The fact that they are not employees, or that he does not retain a right to control them in the manner in which they carry out their work, should not affect the existence of an obligation to prescribe a safe system." (Ibid, at p31).

His Honour stated the working conditions of the operation which had the effect of invoking that principle as being:–
".... that a relationship of proximity existed between Brodribb and the individual worker sufficient to ground a common law duty of care. Subject to the ultimate control of the Commission, Brodribb had an exclusive licence to cut and take away logs from the logging areas. It allocated fellers, sniggers and truck drivers to specified parts in those logging areas; it required them to work together in teams in an intricate process of extracting timber from the forest and delivering it to the sawmill; and it monitored and co–ordinated the operations through its bush boss. While individual fellers, sniggers and truck drivers may have been responsible for their own safety with regard to carrying out their own functions, they had little choice but to rely on the care and skill of Brodribb in the arrangements which it made for the disposition of the work, and on the care and skill of the persons engaged by Brodribb in the execution of the work. The interdependence of the activities carried out in the forest, the need for co–ordination by Brodribb of those activities and the distinct risk of personal injury to those engaged in the operations, called for the prescription and provision of a safe system by Brodribb. Omission to prescribe and provide such a system would expose the workers to an obvious risk of injury." (Ibid, at 30–31).

In the present case it is argued that the same principle applies by reason of the fact that BPA remained in occupation of Stage 2 of the building, and retained control of the co–ordination of the construction operations and control of the site. The evidence relied upon to establish this degree of co–ordination and control is contained in some answers to interrogatories made in an affidavit by Mr SG Thiessen, the Secretary of BPA (exhibit O13), and in evidence given by Messrs Berwick, Chappell and Cooper. The answers say that Mr Berwick was the engineering assistant employed by BPA at the time of the fire, and part of his duties included the co–ordination of contractors engaged on the construction of Stage 2 of the cold store complex, the supervision of the Board's (ie BPA's) day labour force and the inspection of the quality of the finished work. Mr Berwick in evidence said that his role was co–ordination of activities of contractors and ensuring that they followed the construction schedule. The Port Engineer was responsible for the site, and he was the Port Engineer's representative on the site. There was a construction schedule, giving the sequence of work for each contractor, and one of his functions was to co–ordinate the contractors He said that insofar as the standard or quality of their work was concerned, his duty was to monitor them to the extent of endeavouring to ensure that the work overall should be completed to the required standard. Mr Chappell gave evidence that he expected Mr Berwick, as his engineering assistant, to "keep control of the project from a liaison point of view, to ensure that the time scales were met". He also said he would expect Mr Berwick to report to him any serious problems that he perceived, but that he was not expected to exercise control in any way over the employees of the contractors I accept all that evidence. Mr Cooper's position was not materially different from that of Mr Berwick when Cooper was on site.

Overall, I do not accept the argument that BPA in relation to the cold store construction was in an analogous position to that of Brodribb in Steven's case. The two cases are clearly distinguishable on the basic facts. In Stevens v Brodribb (supra) the essence of the relationship between the timber company and the contractors was that the contractors were carrying out various parts of the timber company's business operation in much the same way as they would have been if the company had operated with the use of its own employees rather than independent contractors Moreover, it exercised an extensive degree of control, supervision and co–ordination of the whole operation. The relationship of BPA to WS and, so far as is material, OGP, was quite different. It was not carrying on its ordinary business operations with the aid of independent contractors instead of its own employees. Instead, it was a principal who had engaged an independent contractor to perform an operation in which the contractor was the skilled operative, and the principal relied upon the contractor to carry out the agreed work within the terms of the contract. That is to say, the relationship was that of principal and independent contractor in the traditional sense. However, because there was no head contractor, and the principal was carrying out part of the operation with its own day labour force, and was performing the necessary function of co–ordinating the operations of the two independent contractors and its own day labour force, it necessarily had an ongoing role of liaison during the construction. In my opinion, there was not the requisite proximity relationship between BPA, WS and GJ to provide the basis for holding that a duty of care was owed by BPA to GJ in respect of the way in which WS carried out its contract. Therefore, in my view, BPA was not liable to GJ for any damage caused by negligent performance of that contract by WS.

It is true that the SAA Welding Code contained a clause which provided:–

"2.1 ..........................................
where work has been carried out by a contractor, appointment of the Responsible Officer shall be by negotiation between the principal and the contractor."

The Code is, of course, drawn in general terms designed to be applicable with necessary modifications in a great variety of operational circumstances (see above), and so its application to any particular work situation may be somewhat uncertain. This clause speaks about a "contractor" and "principal", and it would seem prima facie to be intended to relate to a principal contractor and sub–contractors It would seem to be difficult to relate that provision to a building owner or proprietor who had engaged a building or construction firm to carry out an entire building contract, in the course of which of course it would be usual for the head contractor to employ sub–contractors The term "principal" would not make sense as applied to the building owner or proprietor, but "principal" and "contractor" could readily be applied to the builder and his sub–contractors In any case, I have already held that failure to appoint a responsible officer was not in itself a negligent omission on the part of WS, or if negligent was not causative. A fortiori, omission by BPA to discuss with WS the appointment of a responsible officer would not be a negligent omission. I so hold.

The other substantial allegation is that BPA failed to provide any or any adequate means by which a fire in the roof void could be fought, controlled and extinguished. Mr Berwick's evidence was that some six or seven days before the fire he visited the site and asked WS's senior employees, Messrs Chilcott and Whyman if they were providing fire extinguishers on the site. They told him that they had fire extinguishers in their workshops at Devonport, and that they would bring them over to the site. Mr Berwick also instructed his foreman, Mr Clark, to place a fire extinguisher in the roof void in order to cover any operations carried out by BPA's own people. Mr Cooper gave evidence that Mr Berwick asked him also to see that a fire extinguisher was placed in Room 4, but when it was put in place it was thought more useful to put it in Room 5, which was done. It may be that this was the extinguisher which was present there when the fire took place. The evidence indicates that Berwick did not follow up the undertaking by Chilcott and Whyman to have extinguishers placed in the roof void, and none was put there. In my opinion, the contractual arrangements under which BPA was acting as overall co–ordinator of work on the site placed upon it a co–ordinate responsibility with WS, and a duty of care to GJ, and (so far as is material) to OGP, to see that adequate fire fighting equipment was placed in the roof and it failed to perform that obligation. There is no evidence as to how many fire extinguishers would in the ordinary course (that is, not anticipating such a disastrous fire as was in the event brought about by negligent performance of the welding operation by employees of WS) have been considered adequate, but on the evidence as a whole I infer that two would have been adequate. I make this inference partly from seeing the video in which the two extinguishers kept for the purpose by the associates of Dr Cherry who carried out the experiments in which it was necessary to extinguish the fire which they were able to start. Two extinguishers coped very effectively with that fire. However, in my opinion, notwithstanding that BPA failed to perform its duty of care to see that adequate fire fighting equipment was installed in the roof void, such failure was not causative of damage resulting from the fire. The fire was so intense and violent within such a short time after it was first seen by Lillico and Anderson that if there had been two extinguishers present, or even more, the probability is very strong that they would have been wholly inadequate to quench the fire. Accordingly, I find there was no effective negligence on the part of BPA in relation to failure to provide adequate fire fighting equipment. BPA's position is different from that of WS in this respect, because I have held that the failure of WS to provide such equipment was negligent when considered in conjunction with its failure to watch the site after welding ceased. BPA should succeed in the claim against it by GJ, based upon this allegation.

To Olympic General Products Pty Ltd in Action No 2913 of 1981

OGP also claims against BPA for failure to provide adequate fire fighting equipment, and for the reasons just expressed, which apply equally in respect OGP's claim, I find in favour of BPA.

The other issues between BPA and OGP have already been dealt with in respect of Action No 2860A of 1982.

Summary of determinations as to liability.

Action No 694 of 1981

(a)General Jones Pty Ltd is entitled to succeed in full against Wildridge and Sinclair Pty Ltd
(b)General Jones Pty Ltd is entitled to succeed in full against Burnie Port Authority.
(c)Burnie Port Authority is entitled to succeed in full against Wildridge and Sinclair Pty Ltd on the third party notices and on the claims for indemnity and for contribution against WS.
(d)Olympic General Products Pty Ltd is entitled to succeed on the third party claims made against it by the defendants.

Action No 2913 of 1981

(a)Olympic General Products Pty Ltd is entitled to succeed in full against Wildridge and Sinclair Pty Ltd
(b)Burnie Port Authority is entitled to succeed as defendant against Olympic General Products Pty Ltd, and in full against Wildridge and Sinclair Pty Ltd on the claim for indemnity and/or contribution.

Action No 2860A of 1982

(a)Burnie Port Authority is entitled to succeed in full against Wildridge and Sinclair Pty Ltd
(b)Olympic General Products Pty Ltd is entitled to succeed as defendant against Burnie Port Authority; and on the claims for indemnity and for contribution as between it and Wildridge and Sinclair Pty Ltd.