WORKCOVER AUTHORITY OF NSW (INSPECTOR GLASS) V KELLOGG (AUST) PTY LTD.

Prosecution under s15(1) of the Occupational Health & Safety Act 1983.

On the applications of Frank Edward Glass, as an Inspector of the WorkCover Authority of New South Wales ("the prosecutor"), Kellogg (Aust) Pty Limited ("the defendant") was summonsed to appear before the Court to answer two charges that it committed an offence under s15(1) of the Occupational Health and Safety Act 1983 ("the Act"). Both charges concerned an accident which occurred on 26 February 1996 at Stephen Road, Botany. The defendant pleaded not guilty to both charges and consented to the prosecutions being heard together.

By summons filed on 27 August 1997 in Matter No. IRC 4732 of 1997 the defendant is charged:

On 26 February 1996 at Stephen Road, Botany in the State of New South Wales, the defendant, being an employer, DID FAIL to ensure the health safety and welfare at work of all its employees in particular, Salih Husakovic and Ljubisav Jovanovic in failing to provide and maintain systems of work, to wit, systems of operation of the rotary cookers, that were safe and without risk to health, contrary to Section 15(1) of the Occupational Health and Safety Act, 1983, in such case made and provided.

(a) failing to provide and maintain systems of removal of spoiled product from the rotary cookers.

(b) failing to provide and maintain systems of dumping spoiled product from the rotary cookers.

(c) failing to provide and maintain systems of checking internal cooking vessel pressures before cookers are opened.

(d) as a result of he (sic) said failures, Ljubisav Jovanovic and Salih Husakovic suffered burns to their bodies.

By summons filed on 27 August 1997 in Matter No. IRC 4733 of 1997 the defendant is charged:

On 26 February 1996 at Stephen Road, Botany, in the State of New South Wales, the defendant, being an employer, DID FAIL to ensure the health safety and welfare at work of all its employees, in particular Salih Husakovic and Ljubisav Jovanovic in failing to maintain plant, to wit, the rotary cookers, to ensure the health and safety of employees at work, contrary to Section 15(1) of the Occupational Health and Safety Act, 1983, in such case made and provided.

(a) failing to maintain the Fisher automatic steam control actuator.

(b) as a result of the said failure, Ljubisav Jovanovic and Salih Husakovic suffered burns to their bodies.

The immediate events leading to this fatal incident are not in dispute. On 26 February 1996 there was an accident at the defendant's factory on the corner of Stephen Street and Swinbourne Street, Botany, New South Wales. Salih Husakovic suffered severe burns to his body and his workmate Ljubisav Jovanovic suffered burns to his hands. The accident happened at approximately 1.24 am. Mr Husakovic and Mr Jovanovic were employed by the defendant on the 3 pm to 11 pm shift but were on overtime on the night of the accident. They were working on a production line (called the "combination line") which contained a line of six large industrial cookers producing breakfast cereals. Each cooker consisted of a revolving drum which was charged with grain and flavouring and into which steam was introduced. In the days prior to the accident the product was Special K which did not involve all the cookers in the line being in simultaneous production. The accident happened on cooker number 4.

Cooker number 4 had been charged with grain and cooked some five days prior to the accident. It had resulted in a "bad cook". This means that the product was unusable. The decision was taken to shut off the steam to the cooker and to leave it rotating until the contents had been completely dried when they would be easy to remove and the cooker could be used again for production. Cooker number 2 had also been charged and cooked at the same time and also resulted in a bad cook and had been left in the same way. At the end of their shift on 25 February, Mr Husakovic and Mr Jovanovic were instructed to discharge the contents of cookers 2 and 4 and, having discharged cooker 2 with relative ease, found they were struggling with the lid of cooker 4. With difficulty Mr Husakovic finally loosened the nut on the lid of cooker number 4 with the result that there was an explosion of product from within the cooker with such force that it blew out the grating on the walkway situated above the cooker and struck and adhered to the ceiling some 4.5 metres above the cooker. Mr Husakovic luckily managed to avoid the full force of the molten product but some landed on him and he was badly burned. Mr Jovanovic sustained burns to his hands when he went to Mr Husakovic's rescue and tried to remove the latter's clothing.

Mr P Skinner, of counsel, who appeared for the prosecutor, called as witnesses: Inspector Glass, who made an inspection some few hours after the accident; Salih Husakovic, who had been employed by the defendant at the time of the accident for approximately 22 years; Ljubisav Jovanovic, who had been employed by the defendant at the time of the accident for 13 years; Raul Camejo, an employee of the defendant of long standing who worked on the combination line at the time of the accident; Peter Nowak, who was employed by the defendant on the combination line at the time of the accident; Adam Zolnierczyk, a machine operator who commenced work on the combination line in 1985; and Craig Lloyd Allen, who at the time of the accident was employed by the defendant as a business unit resources trainer. Mr Allen relieved facilitators and was involved in the design, development, co-ordination and planning of training.

Mr N Carson, solicitor, who appeared for the defendant, called: Neville John Rawding, who was, at the time of the accident, the Operations Leader on the combination line; and Associate Professor Michael Wootten, who was an Associate Professor in the Department of Food Science and Technology at the University of New South Wales.

On 25 May 1999, after final submissions, a notice of motion was filed by the defendant seeking, in essence, to re-open the trial to allow the admission of further evidence. The notice of motion was initially listed on 4 June 1999. On that occasion the prosecutor opposed the motion on jurisdictional and other grounds. However, after being referred to the judgment in R v Delon (1992) 29 NSWLR 29, when the matter was relisted on 1 July, the prosecutor did not oppose the motion.

The decision of the NSW Court of Criminal Appeal in Delon provides authority for the proposition that the Court possesses a discretion to permit the introduction of evidence at any stage in the trial before verdict, if the introduction of the evidence is necessary to avoid a miscarriage of justice (at 35). Priestley JA stated (at 36):

From the point of view of an accused however, if for instance evidence previously unavailable and of a decisive kind came to hand before verdict, it is hard to see there would be a miscarriage by its being allowed to be put to the jury even at that stage. If this required further addresses by counsel and further summing-up by the judge, that would be a more appropriate course to follow than to leave it to the Court of Criminal Appeal later to set aside the verdict and order a new trial.

In a case of a prosecution under the Act, where the defendant seeks to re-open the trial to introduce evidence which is relevant to the question of whether the Act has been contravened or whether a defence has been established under s53, the Court has a discretion to do so. It may be appropriate to exercise that discretion in a range of circumstances, such as where the evidence was previously unavailable or where the defendant could not have reasonably anticipated that it would be necessary to call the evidence. It is unnecessary in this case to consider the position where the prosecution seeks to call further evidence after the final submissions and where additional considerations may arise: Dryburgh v R [1961] HCA 54; (1961) 105 CLR 532.

On 1 July, in an ex tempore judgment, the Court granted the defendant's notice of motion. The evidence was relevant and, in my view, it was necessary to admit the evidence to avoid a miscarriage of justice. In the result, on 5 July, the Court heard evidence from: Annette Mary Waterworth, who was the defendant's Occupational Health and Safety Manager; Ross Bannerman, who in 1985 was an engineer working on the commissioning of the combination line of cookers; and Adrian Vanslambrouck, who was, in 1985, an Electrical Trainer working on Mr Bannerman's commissioning team.

Apart from hearing the abovementioned witnesses give evidence, I have had the opportunity to re-read the transcripts of their evidence. A number of exhibits were introduced, all of which I have had the opportunity to study.

I find that the following salient facts are proven to the criminal standard of proof, i.e. beyond reasonable doubt.

To convert grain into breakfast cereal it is necessary to cook it with flavourings. This is done by use of the revolving cookers I have described. The cooking is effected by the introduction of steam. I was not told of any other means of heating the cooker. The steam is forced into the cooker under pressure and is exhausted from it continuously, but in such a way as to maintain the pressure, by means of pipes whose inlets are situated near the top of the space within the cooker. The fact that the grain cooks under pressure is evidenced by the heavy lid and heavy securing bars for it. The steam enters the cooker through a slotted fixed tube in the central axis of the revolving drum (which is horizontal) and the ingress of steam is controlled by means of valves situated outside. The production process for Special K in any of the cookers in the combination line is that the cooker is charged with 350 kilograms of kibble wheat (possibly with the addition of flavouring - there is a difference between the accounts given of two of the operators, a difference which I need not resolve) and the kibble wheat is steamed for 50 minutes. The steam which has not been absorbed by the wheat in the form of moisture or exhausted by the exhaust pipes is then discharged from the drum, the lid is opened and 1,050 kilograms of rice and 300 kilograms of flavouring are added to the drum. The drum is sealed again and steamed for a further 50 minutes. Again, the pressurised steam is discharged from the drum which may then be safely opened and the product removed.

Both rice and kibble wheat expands when cooked. This fact is evident to all who have used such substances in domestic cooking. I have had the additional advantage of evidence of a memorandum forwarded by Mr Hernando Prieto, the defendant's Pilot Plant Technician of Research and Development, and a facsimile transmission from Mr Ron Berry, Research and Development Manager of Rice Growers Co-operative Limited to Inspector Glass. Mr Prieto stated that rice and kibble wheat cooked together for 1 hour at 100 degrees at a pressure of 140 Kpa will expand by 36.6%. Mr Berry's evidence was that rice when cooked with water to an acceptable level increases by 36% in volume. There was no evidence before me as to what might happen to a cooker which was full of rice and kibble wheat, subjected to steaming for 6 hours, left without steam for several days and then steamed again for some 22 hours. Neither was I provided with any evidence as to what quantity of steam and/or water one might find in a cooker subject to such additional steaming.

From time to time when cooking breakfast cereals, including Special K, a "bad cook" will occur. When that occurs nowadays, i.e. since the accident, the procedure is to dump the contents of the cooker immediately. That was the evidence of Mr Zolnierczyk and it is borne out by the re-written manual produced after the accident.

Prior to the accident a bad cook had occurred on 21 February 1996 in both cooker number 2 and cooker number 4. Mr Camejo had been instructed to leave the contents in the cookers (without steam but revolving) until the production run of Special K had ceased a few days later. This was to save discharging the cookers into the common discharge auger which ran beneath the cookers and which would have to be cleaned after the bad cooks had passed through it and before it could be used for good products. The contents of cookers 2 and 4 were, therefore, left for some four and a half to five days. Messrs Husakovic, Jovanovic, Allen and Nowak all gave evidence that such a long delay had never occurred before. Mr Camejo gave conflicting evidence saying at one point that such a delay was usual but later that it had never occurred. To the extent that it is necessary to resolve this issue I find that never before had a bad cook been left in a cooker for so long. The intention in keeping the bad cooks revolving without steam was to render the contents dry, cool, loose and powdery (apart from the occasional lump) and relatively easy to remove when the time came to discharge the contents. This proved subsequently to be the case with cooker number 2. This process is distinguishable from that described by Mr Zolnierczyk whereby at earlier times water was added to the cooker and boiled for a period of time.

It appears that the reason, or at least one of the reasons, why the load in cooker number 4 was a bad cook was that the amount of grain with which it had been charged exceeded the specified load I have mentioned above. This was the evidence of Mr Husakovic, and Mr Carson effectively conceded that this was the case (transcript page 51, lines 26-27). Since the load was excessive the problem could have been resolved by discharging the load before it was steamed but the Court was informed that accidents of excessive loading did occur from time to time and were inadvertently cooked. Mr Skinner did not suggest that this constituted a breach of s15.

Mr Jovanovic firstly started to open cooker number 4. It was a "little bit tense" to commence. After about four or five turns of the bolt to open the cooker lid Mr Jovanovic observed a substance flowing from the partially open lid. Subsequently Mr Husakovic observed this substance. It was variously described by the operators as "watery muddy stuff" and "muddy like tar" and being of a "really dark colour". Both operators described this situation as unusual - they had not experienced such an event previously.

Mr Husakovic subsequently continued to open the lid after the flow of this substance ceased. When he managed to loosen the lid on cooker 4 the explosion consisted of product of a porridge-like consistency which was so hot that it caused him extensive burns resulting in 4 weeks in hospital. This was despite Mr Jovanovic's presence of mind in turning the cold water hose on Mr Husakovic after having attempted to remove Mr Husakovic's clothing. The product was so hot that it burned Mr Jovanovic's hands in trying to get Mr Husakovic's clothes from his body. From this evidence I conclude that the contents of cooker number 4 had been subjected to live steam for some time.

Before the incident in 1996, there had been at least one previous incident involving material erupting from a cooker. In his evidence, Mr Zolnierczyk described an incident which was said to have occurred around 1985, 1986 or 1987. Mr Zolnierczyk said that the procedure employed at that time in the case of a bad cook was to leave the material in the cooker with water added to produce a soup-like substance. The operators would then remove the lid via a mechanical device operated from the control room. Mr Zolnierczyk testified that on one occasion when he pressed the button to remove the lid the material inside the cooker erupted with considerable force reaching the ceiling of the factory. The eruption resulted in a majority of the material being extruded from the cooker. Mr Zolnierczyk described the eruption as being like "Vesuvius".

The defendant introduced additional evidence in an attempt to shed further light on this incident. Ms Waterworth gave evidence that she searched the company records and could find no reference to any previous incident of the nature described by Mr Zolnierczyk. Ms Waterworth made inquiries of current and former employees of the defendant, none of whom could remember any such incident aside from Mr Bannerman and Mr Vanslambrouck. Mr Bannerman, who worked on the commissioning of the combination line cookers during 1985, gave evidence that a lot of problems were experienced during the commissioning of the cookers. He recalled one incident in which a command was sent to the wrong cooker resulting in the de-lidding of the cooker during the cook phase whilst the cooker was under pressure. This caused the lid to explode off the cooker. Mr Bannerman indicated that the explosion was not due to a bad cook, but from the lid being removed in error and that steps had been subsequently taken to prevent any recurrence. The evidence of Mr Bannerman was supported to some extent by the evidence of Mr Vanslambrouck, although he did not see the incident in question and was unable to give any insight concerning why the accident occurred.

In relation to the previous incident or incidents, Mr Skinner submitted that the evidence did not establish that Mr Zolnierczyk, Mr Bannerman and Mr Vanslambrouck were referring to the same incident. The nature of the incidents described by Mr Zolnierczyk was fundamentally different from that described by Mr Bannerman and Mr Vanslambrouck and Mr Skinner submitted that "they must be referring to a completely different incident". Further, the absence of any record of the incident described by Mr Zolnierczyk did not establish that it did not occur. Rather, the failure to record any such incident illustrated the inadequacy of the recording mechanisms employed by the defendant at that time. Mr Skinner also pointed out that it was never put to Mr Zolnierczyk either that the event did not occur or that he had mistaken the nature of the event. He submitted that whilst the rule in Browne v Dunn (1894) 6 R(HL) 67 does not apply strictly in criminal proceedings, the failure to confront Mr Zolnierczyk should be considered when attaching weight to his evidence.

Mr Carson submitted that the additional evidence was adduced primarily to support the submission that the defendant had established a defence under s53 in that there was no basis for the defendant to foresee or apprehend that an event, such as happened, could have occurred by reason of leaving the material in the cooker for five days. He submitted that it is clear that the incident which Mr Zolnierczyk was referring to is the same incident as that described by Mr Bannerman and Mr Vanslambrouck. The evidence of Mr Bannerman and Mr Vanslambrouck indicated that the extrusion incident was the result of an error which occurred during the commissioning process and was associated with other problems which occurred at that time. It was submitted that there were no other incidents outside the commissioning period and that the defendant could not have foreseen the possibility of the event occurring.

I find it unnecessary to resolve the differences of recollection in relation to this incident or incidents. What is beyond doubt, and hardly needed an explosive incident to demonstrate, was that these pressurised cookers were extremely dangerous if the lid was removed or loosened whilst there was pressure within. The incident also illustrates something that was evident throughout the hearing before me, namely that the defendant's record keeping was poor. There were no reports produced of the incident as described by Mr Zolnierczyk or by Mr Bannerman. The log book for the period surrounding the accident has been lost and faults which were repaired were unknown to Mr Rawding who was supposed to have access to the paperwork.

The cookers had pressure gauges situated on the machine. Pressure was also readable in a digital form from a computer screen situated relatively nearby. The pressure gauges recorded steam or air pressure, but they were not capable of recording the pressure of solids such as product within the cylinder. Although there is some suggestion in the evidence that the pressure gauges recorded the pressure of the steam passing through the pipe and valve into the cooker, I find that the gauges actually measured steam pressure within the cooker. It seems to me that for the productive process it was essential to know the Kpa within the cooker so as to ensure the appropriate pressure for cooking the product. Whilst the pressure within the steam pipes may have been of interest to relevant persons to ensure that the pressure in the pipes was sufficient and appropriate for supply in the productive process and not so high as to threaten the integrity of the pipework, valves and equipment, I was not told that that information was necessary for the cooking process. Hence the Kpa recorded on exhibit 10, a print-out from the computer, appears to me to be the pressure within the cooking chamber which was the information required for production. This, it appears, was the evidence of Inspector Glass who stated:

... If there was steam pressure within the cooker the gauge would show the pressure of what one would call air pressure the steam (sic), and it would be indicated on the gauge and also indicated on the print-out.

I accept that the pressure shown on the gauges and computer were not capable of recording the pressure of the solids in the cooker but only of air or steam within it, but if the gauges and computer showed the existence of pressure within the vessel there seems to me to be no reason to suppose that all the contents thereof, i.e. rice, kibble wheat, water and steam, were not under compression.

The steam valve to cooker 4 was tested on 28 February 1996 and found to be leaking. Mr Rawding concedes that this was so but gave evidence that he did not know of the leak prior to its discovery on investigation after the accident. The existence of the leak is made plain by the report of a contractor, Optimum Control Pty Limited, dated 14 May 1996. The valve was subsequently repaired. Since the accident an additional manual valve has been inserted into the steam line. Though Mr Rawding was unaware of the steam leak before the accident, Mr Husakovic gave evidence that he knew of it and also that there was a leak at the back part of the cooker. Mr Jovanovic also gave evidence that he too knew that the valve was leaking prior to the accident and that it had been reported. Mr Camejo gave evidence that the leaking valves had been reported 6 months before the accident in relation to cooker number 2 and cooker number 4. He gave evidence that the valves were reconditioned but not until after the accident. In the absence of the log book to provide any indication to the contrary, this evidence would tend to indicate that a steam leak occurred in cooker number 4 and that such steam leak was not repaired prior to the accident.

However, documentary evidence from Optimum shows that they serviced the valves prior to 30 September 1995 and calibrated them prior to 25 January 1996. It is quite possible that any leaks were remedied on one or other of these occasions and that the recollection of the men may not be accurate in their memory of a leak or leaks persisting after 25 January. I find I have reasonable doubts as to whether there was a continuing steam leak after Optimum had attended in January 1996.

Exhibit 10 has assumed some importance in this case. It is a print-out from a computer situated near the combination line of cookers. That computer apparently demonstrates (and records) much information about the cookers (although I was shown no other print-outs from it). Exhibit 10 is a graph with its vertical axis marked for pressure measured in Kpa and its horizontal axis marked in time running from the left at 7.00 am on 20 February 1996, running through to the right which is 7.00 am on 1 March 1996. Exhibit 10 had been extracted and prepared by a Mr Wilson, the defendant's Security and Safety Co-ordinator. It is a printout of a screen which could have been accessed on the screen of the computer. It has been referred to as a "trend screen".

The print-out has been marked at the bottom with various letters corresponding to the rise and fall in pressure during the time of the graph lines (an enlarged version of exhibit 10, with date and time entries corresponding to these letters, was exhibit B in the proceedings). These markings are artefacts added for the forensic purposes of this case and were not on the screen or the printout as originally generated. A series of vertical lines at A and B show some five cooks on cooker 4 which ended when Mr Camejo shut the cooker down after releasing the steam therefrom for some 45 to 60 minutes on 20 February 1996. He had realised this was a bad cook. At 9.23 pm on 21 February pressure had increased in the cooker for a period of just short of 6 hours and was then released. No-one has been able to offer an explanation for this peak in the graph. Mr Carson, in final submissions, put that Mr Rawding did not think that this was a leak. I have checked his evidence again and it does not appear to me that Mr Rawding was excluding the possibility of a leak; it was simply that he could not explain the peak recorded on the graph. Since the graph records steam pressure the conclusion I should draw is that steam had leaked into the cooker for a period of nearly 6 hours. That it was a leak rather than, as Mr Carson suggested as a possibility, a deliberate tampering with the valve, seems to me to be supported by the slow rise in pressure over several hours illustrated by a jagged, but roughly diagonal, line in contrast to the sudden rise in pressure (illustrated by a near vertical line) on the many other occasions on this graph when the cooker has been charged with steam by purposely opening the steam valve. What caused the sudden discharge of pressure at letter D on the graph also remains a mystery. The line is as close to vertical as the other controlled discharges of steam.

At E to F on the graph is shown a rise in pressure beginning at 3.08 am on 25 February 1996 and rising in a jagged line inclined at about 70 degrees (although varying in its course) from the horizontal and persisting until 1.24 am on 26 February 1996. On this occasion there is no doubt about the sudden discharge of pressure which was caused by Mr Husakovic loosening the lid. Again, despite Mr Carson's attempts to persuade me of other possibilities, I conclude that the rise in pressure to some 74 Kpa over the period in excess of 22 hours was caused by steam entering the cooker. In the absence of any other credible explanation and consistent with the shape of the "curve" on the graph, I conclude, beyond reasonable doubt, that this ingress of steam was caused by the leaking valve subsequently discovered on 28 February 1996.

The graph at the remaining letters between G and L show other events subsequent to the accident, including the testing of the valve and the demonstration which proved that it was leaking. I need not further consider these events.

The incidents at C to D and E to F demonstrate that the bad cook in cooker number 4, which had, by 21 February, already undergone a full cooking process, was subject to a further 28 hours of steam with an interval of some days between the first 6 hours and the remaining 22 hours. Mr Rawding gave evidence that that information was available if the screen had been inspected, but he gave evidence that the operators were not told to use this trend screen. Mr Jovanovic and Mr Camejo also confirmed that the operators never used that screen.

I conclude, beyond all reasonable doubt, that at 1.24 am on 26 February 1996 the load in cooker number 4 consisted of grain, flavouring, water and steam at great pressure and at a considerable temperature. That conclusion would have been self-evident to any operator or manager who looked at the trend screen on the computer from which the print-out, exhibit 10, was subsequently taken.

Mr Husakovic and Mr Jovanovic each gave evidence that they had checked the pressure gauges on cooker number 4 and the equivalent information found on the computer, but not the trend screen on the computer. They gave evidence that the pressure gauges read zero pressure. I do not doubt that they are telling the truth as they believe it to be. However, I am unable to accept their evidence on this point.

I accept that they checked the gauges on cooker number 2 before discharging it. However, I have concluded that having successfully discharged cooker number 2, they did not check the pressure gauges on cooker number 4 before wrestling with the lid. Having examined that the valves were appropriately positioned and having successfully emptied cooker number 2, I consider Mr Jovanovic then, without further checks, immediately proceeded to loosen the lid and after which the "tar" like substance flowed from the lid opening.

The operators gave unchallenged evidence that there was no steam coming from the partially opened cooker. The absence of steam emission from the partially opened cooker (in which there was a steam leak) was not explained in the evidence in the proceedings. However, this fact does not necessarily support the evidence of the operators that the gauges were checked and showed a zero reading. For example, the manual subsequently produced by the defendant refers to the prospect of a "steam pocket" in a cooker (this prospect giving rise to a requirement in the new procedures applying in relation to bad cooks that the cooker be rotated prior to opening).

The operators gave evidence that the pressure gauges and computer screen were checked after the initial opening of the cooker lid and the excretion of the tar like substance. This evidence was not challenged in cross-examination.

However, exhibit 10 clearly demonstrates the existence of steam pressure at the time cooker number 4 was opened. The pressure reading on the computer trend screen is, as I understand it, simply a different display of the same information which is provided on the pressure gauges next to the cooker itself at various points in time. The tests by the maintenance contractors, Optimum, showed steam pressure was generated by the leaking valve, as evidenced on the computer screen in exhibit 10 at letters J to K. No-one has suggested that the mechanical gauges were defective or that they showed (or were capable of showing) zero pressure whilst the computer showed significant pressure. Since the computer screen at letter F demonstrated pressure of 74 Kpa immediately before Mr Husakovic loosened the lid, it must follow that the pressure gauges on the cooker also demonstrated a similar pressure (subject to any slight variations of calibration).

On the occasion of the accident Mr Husakovic and Mr Jovanovic were carrying out a very familiar task at the end of a long shift in the early hours of the morning. The operators, such as Mr Husakovic and Mr Jovanovic, effectively worked without supervision and were encouraged to do so. The facilitators had a little knowledge of the productive processes and were certainly not in a position to advise experienced operators. Specifically, they were not, and did not, act as supervisors. The operators had to rely on their own resources and were instructed to do so. They ultimately decided if outside assistance may be required either through a facilitator or by reference to maintenance personnel. They were able to report faults but had no control over when, or indeed if, the faults were rectified.

The submissions of the parties were made orally. I have re-read the transcript of the submissions and provide a brief summary of them (although I note that I have dealt with aspects of the submissions of the prosecutor and defendant elsewhere in this judgment).

Mr Skinner alleged two distinct failings on the part of the defendant. It was submitted that the defendant failed to provide a safe plant because the equipment employed was defective in that the steam valve attached to the cooker was leaking at the time of the accident. Mr Skinner alleged that the leaking steam valve was an operative cause of the accident because it permitted the material inside the cooker to continue to cook and expand during the five days it remained in the cooker. It was submitted that the defendant did not have in place an adequate system for the regular maintenance of the steam valve, nor for reporting leaks or defects in the valve or logging previous untoward incidents. In any case, it was put that the leak was known prior to the accident and steps should have been taken to do something about it. Mr Skinner submitted that the risk to safety thereby created was readily foreseeable to anyone with common experience of the characteristics and behaviour of rice products cooked by steam over a lengthy period in an enclosed vessel. In this case, the presence of this kind of risk was foreseeable in the light of the previous occurrence some years earlier.

In addition, Mr Skinner submitted that the defendant had failed to put in place a safe system of work for the removal and dumping of spoiled product from the rotary cookers. The system of leaving the spoiled product in the cooker for up to a week and then requiring the employees to manually remove the lid of the cooker was inherently unsafe. A safe system of removal would allow the spoiled product to be removed without the necessity of a worker standing in the immediate vicinity of the cooker. Furthermore, the defendant failed to provide a volumetric gauge demonstrating the pressure of the material, and not just the steam, inside the cooker. The operators were not instructed to check the pressure of the steam in the cooker prior to opening it.

In regards to the use by the operators of a display on a computer showing the trend of pressure within the cookers (which was described in the evidence as the "trend screen"), Mr Skinner stated:

The ability was there to go and look at the computer screen which when printed out has become in this court exhibit 10 and exhibit D if the operators had known or been instructed to do that, but there is no evidence that indeed they were.

The evidence of Mr Jovanovic on that issue appears at page 59 of the transcript he was asked by me: Did you ever look at a computer screen or other display showing you that sort of display? - this was in the context of questions about a computer gauge, it is obviously what has become exhibit 10 - he said, "Actually we never did that..." The thrust of that evidence is even though that screen was available to be looked at it was not the practice and he did not even know it was there to be looked at and that evidence was effectively confirmed this morning by Mr Rawding that indeed to his knowledge it was not an instruction to go and look at that. If they had done that and understood what it meant in the way we now understand, this court and your Honour has received evidence of what that screen means, the operator would have seen there had been a leakage of steam and they could have done something appropriate other than to knock the lid of the cooker off with a hammer.

It was also submitted that there was also inadequate mechanisms for ensuring proper communication between workers on different shifts and inadequate supervision of the workers. In their work the operators were encouraged to be self directing. Mr Skinner submitted that although the operators were paid a remuneration which to some degree was a reflection of the responsibility they were given, "the evidence is there really wasn't the degree of supervision that probably should have been in place". He emphasised that given the dangerous nature of the cookers, the lack of supervision was a failure to carry out a safe system of work.

It was submitted that the inadequacy of the system operated by the defendant was demonstrated by the improved procedures incorporated in the defendant's operations manual which was revised after the accident.

It was also submitted that at the time of the accident there were no procedure or instructions in place for the disposal of bad cooks. The fact that the company produced such instructions or procedures after the accident is a factor going to mitigation. The fact remained, he contended, that these procedures should have been in place after the earlier accident.

Mr Carson submitted that the accident was a "freak occurrence" which cannot be properly explained by the evidence and that the defendant did have in place a system for the disposal of spoiled product which was not dangerous. In the first place, Mr Carson submitted that Messrs Jovanovic and Husakovic caused the accident by attempting to remove the lid from cooker number 4 and had acted in a manner which was unpredictable and against which the defendant could not have been expected to have taken precautions. They were, it was said, "authors of their own misfortune". Messrs Jovanovic and Husakovic were dealing with a situation which was outside their experience and should have sought help from the facilitator. Had this occurred, the lid would have been left in place until expert assistance could have been obtained. Mr Carson emphasised that there was no danger so long as the lid of the cooker was not removed.

Mr Carson argued that there was no firm evidence presented as to what caused the increase in the pressure in cooker number 4 which resulted in the accident. He argued that it was possible that the increase in pressure in the cooker had resulted from the leaking steam valve, but it was equally possible that it had resulted from an unknown manual intervention. There was reasonable doubt concerning whether the pressure in the cooker was caused by the leaking valve. Furthermore, evidence was given by Messrs Jovanovic and Husakovic that they checked the computer and the other gauge prior to opening the cooker and that the pressure was nil. This evidence was in conflict, he submitted, with the readout from the computer contained in exhibit 10.

It was submitted that, in the absence of a relevant finding as to what had occurred, the prosecution has failed to prove its case. In this regard, Mr Carson relied upon a line of authority including McMartin v The Broken Hill Proprietary Company Limited (unreported, Grove J, Matter No. 20002 of 1986 , 5 February 1988), Cullen v State Rail Authority of NSW (1989) 31 IR 207 and State Rail Authority of NSW v Dawson (1990) 37 IR 110. Further, Mr Carson contended that the regime in place for the maintenance of the steam valves, including annual overhaul and recalibration and regular inspection, was an appropriate and sufficient system. He submitted that it would be unreal to require more frequent testing of the valves and equipment than was already undertaken by the defendant.

Mr Carson further submitted that a safe system was in place prior to the accident. The essential submission advanced in this regard was that the prosecution agreed that the system now in place was safe. He pointed out that the only changes implemented following the accident involved the production of a manual to instruct workers as to what to do and to remove bad cooks from the cooker immediately rather than waiting for up to a week. It was also contended that there is no causal connection between an alleged failure to have a manual of the nature produced after the incident because the procedures set out in the manual were the same as those in place before the incident. The operators had carried out the manual or automated operations for some eleven years or so on a number of times per shift and so did not require such a manual.

It was submitted that there can be no causal connection between an alleged failure to have a manual in place and the incident because the procedures were the same as they were before. Either the procedure is safe or not. Leaving the spoiled product in the cooker was also not causally connected to the incident. Mr Carson argued that it was perfectly safe to leave the spoiled product in the cooker rotating slowly, so long as no water or steam was present. That aspect of the system did not cause the accident. The absence of a volumetric gauge was not critical.

Mr Carson also submitted that there was no danger until the lid to the cooker was opened. He submitted that the operators had made a wrong decision as to the question of the situation with cooker number 4 having stabilised before proceeding to open the lid. Instead, they should have called for a facilitator rather than proceeding with a situation they had never encountered before. There was no set procedure as to what a facilitator would have done in the situation but it was submitted that "there was actual evidence of what could have happened".

In regards to the use of the computer screen and the trend graphs (exhibit B) Mr Carson submitted:

It is not suggested, I think by anyone, that it was something which was a day to day tool used by the operators. They had a separate computer programme, which showed the pressure at the time - it was referred to as in digital.

In the alternative, Mr Carson submitted that the defendant had made out a defence under s53 of the Act, namely, that it was not reasonably practicable to comply with the Act or the commission of the offence was due to causes over which the defendant had no control and against the happening of which it was impracticable to make provision. He emphasised that the accident resulted from Messrs Jovanovic and Husakovic acting in a manner which was unpredictable in circumstances where there was no foreseeability of danger. What had occurred was completely untoward and the cause of which was not decisively revealed by the evidence. These circumstances, it was submitted, establish a defence under s53.

(1) Every employer shall ensure the health, safety and welfare at work of all the employer's employees.

(2) Without prejudice to the generality of subsection (1), an employer contravenes that subsection if the employer fails -

(a) to provide or maintain plant and systems of work that are safe and without risks to health;

(b) to make arrangements for ensuring safety and absence of risks to health in connection with the use, handling, storage or transport of plant and substances;

(c) to provide such information, instruction, training and supervision as may be necessary to ensure the health and safety at work of the employer's employees;

(d) as regards any place of work under the employer's control -

(i) to maintain it in a condition that is safe and without risks to health; or

(ii) to provide or maintain means of access to and egress from it that are safe and without any such risks;

(e) to provide or maintain a working environment for the employer's employees that is safe and without risks to health and adequate as regards facilities for their welfare at work; or

(f) to take such steps as are necessary to make available in connection with the use of any plant or substance at the place of work adequate information -

(i) about the use for which the plant is designed and about any conditions necessary to ensure that, when put to that use, the plant will be safe and without risks to health; or

(ii) about any research, or the results of any relevant tests which have been carried out, on or in connection with the substance and about any conditions necessary to ensure that the substance will be safe and without risks to health when properly used.

(3) For the purposes of this section, any plant or substance is not to be regarded as properly used by a person where it is used without regard to any relevant information or advice relating to its use which has been made available by the person's employer.

Maximum penalty: 5,000 penalty units in the case of a corporation or 500 penalty units in any other case.

(4) If in proceedings against a person for an offence against this section the court is not satisfied that the person contravened this section but is satisfied that the act or omission concerned constituted a contravention of section 16, the court may convict the person of an offence against that section.

Without prejudice to the generality of the duty cast on an employer by sub-section (1), sub-section (2) of s15 then sets out the indicia of how an employer contravenes sub-section (1) by elaborating the general duties imposed by sub-section (1). Section 15(1) establishes a far reaching obligation on an employer and imposes a duty in strict terms with s15(2) spelling out the heads or particulars of that absolute duty but without in any way cutting down its rigour: Dawson at 113.

Since s15(1) imposes an obligation of strict liability it is not to be limited to risks which are reasonably foreseeable. As was stated in Drake Personnel Ltd v WorkCover Authority of NSW (Inspector James Swee Ch'ng) (unreported, Wright J, President, Walton J, Vice-President and Peterson J, 12 August 1999 at 38-39):

We consider that this submission is misconceived. Firstly, the submission focuses on the foreseeability of the alleged risk to safety. The authorities cited above establish that in order to demonstrate a breach of s15(1) the prosecutor is required to prove the existence of a failure on the part of the employer which is causally connected with the risk to safety. There can be no failure on the part of the employer in not taking steps to preclude a risk which was impossible to anticipate: see Dubavo. However, the appellant goes further and submits that no breach of s15(1) is committed where the employer does not know of the hazard and could not reasonably know about it.

The concept of "reasonable foreseeability" is not, in our view, apt to be applied in relation to the duties owed under the OH&S Act. The duties imposed by the Act are not merely duties to act as a reasonable or prudent person would in the same circumstances: see Carrington Slipways Pty Limited v Callaghan (1985) 11 IR 467 at 469. Under s15(1) the obligation of the employer is "to ensure" the health, safety and welfare of employees at work. There is no warrant for limiting the detriments to safety contemplated by that provision, to those which are reasonably foreseeable. Whilst employers are not liable for risks to safety which are merely speculative or unduly remote (see Kirkby v A & M I Hanson Pty Ltd (1994) 55 IR 40 at 50), the terms of s15(1) specify that the obligation under that section is a strict or absolute liability to ensure that employees are not exposed to risks to health or safety. It is inappropriate to seek to substitute a different test for that specified in s15(1).

As previously mentioned, the defendant raised the question of causation. Many decisions of this Court and its predecessors have demonstrated that the mere occurrence of an accident causing injury to an employee is not in itself sufficient to establish the commission of an offence under s15(1). It is necessary to establish both a relevant "failure" on the part of the employer and a "causal nexus" between the conduct of the defendant and the consequent risk to the health, safety and welfare of its employees: see McMartin; Drake at 34; Cullen at 209; Dawson at 120-121; Haynes at 156-157; and Kirkby at 49-50. There must be a causal connection between the alleged conduct of the defendant and the alleged risk, that is, the alleged failure must cause the detriment to the safety of the person concerned: WorkCover Authority of NSW v Maitland City Council (1998) 83 IR 362 at 377.

However, it is not necessary to demonstrate a causal connection between conduct of the defendant and the precise circumstances of the accident which gave rise to the prosecution. Rather, the causal connection must be between that conduct and the risk to safety. In Drake at 39-40 the Full Bench stated:

The general duties created by the OH&S Act are directed at obviating "risks" to the health, safety and welfare of persons in the workplace: see Haynes v C I & D Manufacturing Pty Ltd [1994] NSWIRComm 230; (1995) 60 IR 149 at 158-159. The occurrence of an accident and the sustaining of injuries by an employee will certainly represent relevant evidence of the existence of a risk to the health and safety of employees and the seriousness of that risk. However, it is not the accident itself which constitutes the offence, but rather the failure of the employer to ensure that its employees are not exposed to risks while at work.

This consideration is carefully explained in the decision of the Full Court of the then Industrial Court of NSW in Haynes. Commenting on the reasoning of the trial judge, their Honours said (at 156-157):

The commencing point of her Honour's consideration of the evidence was the uncontroverted fact that the death of Mr Chen was caused by the front end loader hitting him as it came to a stop over the pit in which he was working removing the slurry. Her Honour then, correctly in our view, commented that "the occurrence of the accident itself does not however establish the commission of the alleged offence": see Dawson (at 120-121) and McMartin v The Broken Hill Proprietary Company Ltd (unreported, Grove J, 5 February 1988, pp 7-8). But, her Honour then concluded, concerning the case against Manufacturing, that "(i)t needs to be established that some action taken by (Manufacturing), or the failure to take some action, was causally connected with the front end loader hitting Mr Chen. It seems to (her Honour) that to approach the question of the commission of the offence in any other way, would be to remove the requirement placed upon the Prosecution to establish, beyond reasonable doubt, that it was the Defendant which failed to `ensure the health, safety and welfare at work' of Mr Chen and Mr Ding. The prosecution of Industrial was considered by her Honour according to "(m)any of the comments ... made in relation to the prosecution against Manufacturing"; again, a causal connection between some act or omission by Industries and the accident was found to be necessary. Her Honour focused attention on the accident itself, that is the front end loader hitting Mr Chen, as the relevant detriment to safety with which the respondent were charged. We think that approach emphasised too much the accident itself as the relevant detriment to safety of persons on the site. That overemphasis led to a misunderstanding by her Honour of the real facts disclosed by the circumstantial evidence and the proper inferences to be objective drawn therefrom.

Sections 15 and 16 of the OHS Act are both concerned with failures to ensure the health and safety of persons at workplaces in terms inter alia of "risks" thereto; thus, the sections, even absent any actual accident causing death or bodily injury, nevertheless comprehend the commission of an offence where the relevant "detriment to safety" (as spoken of in Dawson and McMartin) is but a risk, or, in other words, where the circumstances are such that an employer's act or omission has created a situation of potential danger to the health and safety of persons at his workplace. The OHS Act, as its long title, indicates, has the prime purpose "(t)o ensure the health, safety and welfare of persons at work" and that state purpose may only reasonably be achieved, it seems to us, by construing the general duties or obligation cast on employers by Div 1 of Pt 3 thereof (which contains ss 15 and 16) as both preventive and remedial in nature, that is, both before and after the occurrence of an actual accident.

We adopt that reasoning as a correct statement of law and principle in relation to s15 of the Act.

Having regard to the evidence and the submissions of counsel for the prosecutor and the defendant, I shall now turn to consider the liability of the defendant under s15(1) of the Act. Initially, I shall deal with those aspects of the prosecutor's contentions which do not, in my view, establish a proper basis for a finding of liability under the section.

I reject Mr Skinner's submission that the defendant examined the valves and equipment with insufficient frequency. The valves had been attended to on two occasions in the six months before the accident. The mere fact that there was a leak subsequently (as I have found) does not render this frequency a breach of s15(1) or (2). In any event, the failure to inspect more frequently was not causative since it is not established that the fault developed or could have been detected prior to 26 February and I doubt whether more frequent inspection would have discovered it. Furthermore, if liability did exist in this respect, I would hold that it was not reasonably practicable, pursuant to s53, to inspect more frequently (the defences available under s53 of the Act will be discussed more fully below).

The defendant's record keeping appears to be gravely at fault on the evidence before the Court. The defendant did not have a reasonable system for recording faults and malfunctions and acting promptly upon them. However, there is a reasonable doubt as to whether this fault was causative of the accident since I am not convinced that the leaking steam valve, noticed by the operators prior to the accident, was not dealt with by one or other of the prior visits of the maintenance contractors.

The prosecutor argued that the defendant had failed in not providing a volumetric pressure gauge. It appears to the Court that the pressure gauge, the functioning of which is recorded in exhibit 10, was demonstrably successful in recording the dangerous pressure prior to the opening of the cooker. No other or further gauge was necessary.

I reject the allegation that it was unsafe to leave product within the cooker for five days. That system would have been safe had not steam leaked into it so as to make the contents hot instead of cool and pressurised instead of being at atmospheric pressure.

I also reject the submission that it was unsafe to use a hammer per se to assist in the removal of the lid. Any technique which successfully loosened the lid would have exposed those nearby to the risk of an explosion of the pressurised contents within.

I turn then to consider a number of the submissions advanced by the defendant as to its liability under s15 of the Act.

I reject the suggestion that there was manual intervention which caused the steam to get into the cooker immediately prior to the accident. I have found as a fact that the steam which entered the cooker came from the leaking valve. I reject, therefore, Mr Carson's submission that there is no explanation of why steam leaked on the two occasions at C to D and E to F on the graph. I accept, however, that there is no explanation of why the leaking valve was apparently intermittent with several days of inactivity between C to D and E to F. I accept also that there is no explanation of how the pressure was discharged at point D. I do not conclude that the absence of explanation is fatal to the prosecution's case. It is unnecessary to have that event explained in order to reach a conclusion as to whether the circumstances about which I can make findings amount to a breach of s15. Mr Carson did not seek to persuade me that there was some possible explanation of the discharge of steam at D which could have any bearing on the subsequent accident or its causes.

The release of the lid on a cooker such as number 4 was hazardous. The cooker was built to contain product and steam at pressure. There was an obvious risk (with or without any previous incidents) of a violent expansion of product and/or steam in the event that the lid was loosened or removed whilst there was product and steam under pressure within. This was a reasonably foreseeable risk although it is not necessary for me so to find in relation to s15. It was incumbent upon the defendant therefore to ensure that the system for removal of the lid was safe.

Steam leaks from valves have been a well known hazard since the invention of the steam engine. They were known on this combination production line. They necessitated maintenance and, I have found, the defendant fulfilled its duty to maintain them. But the fact that steam leaks occurred from time to time necessitated constant vigilance, in particular, to ensure that steam had not escaped into and built up pressure within a cooker without the knowledge of those working in the vicinity. This is particularly so in relation to 'bad cooks', which, by nature of the process of 'drying', might otherwise be treated as having been free of any steam input into a cooker (which had been left rotating to dry the product).

I find that the defendant should have ensured the operation of a system of work which ensured that the operators checked the pressure gauges by the cooker, the pressure gauge screen on the computer and the trend screen on the computer prior to loosening a secured lid on any cooker. Had the operators in question inspected the gauges and the computer prior to the accident they would have realised that the vessel was under pressure and would not have attempted to move the lid. Had they looked at the trend screen, they would have also, in my view, realised that the graph showed a steam leak or, if not, they would have realised from the shape of the graph that something was amiss.

Thus, whilst the defendant cannot be criticised for the fact that the valve leaked steam into cooker number 4 nor, on the evidence, for the suggestion that it had been leaking and had not been repaired, the defendant was guilty of a failure to ensure that the operators checked the pressure gauges, including the trend screen, before attempting to remove the lid.

I reject the suggestion that Mr Husakovic and Mr Jovanovic were at fault in failing to have called for a facilitator. On the evidence, there was no competent advice as to the process which a facilitator could have given them and, in any event, they were unaware of any specific danger. As previously mentioned, operators were self-directing and operated without supervision. It is understandable, therefore, that in the absence of a specific procedure and instructions given (particularly in relation to bad cooks), the operators may attempt to resolve a difficulty with cooker number 4 as part of their ordinary duties.

I further reject Mr Carson's submission that Mr Husakovic and Mr Jovanovic were the authors of their own misfortunes. It is, to say the least, unfortunate that the defendant puts that those two persons who were burned in the course of their service with the defendant were "looking for a scapegoat". They sought to blame no-one but tried only to present to me the truth as they saw it. It was WorkCover which brought this prosecution pursuant to their statutory duty. In my view, there was insufficient material before the Court to support the submission advanced.

I find that the duty under s15(1) translates in this case to a duty on the defendant to ensure that where the lid on a cooker has been secured so that any pressure within it cannot escape through the lid, then the lid should not be loosened or removed unless and until it has been ascertained that there is no pressure within by checking the gauges and the computer screen including the trend screen. This is an essential aspect of the duty to provide a safe system to remove spoiled product from the cookers which is at the heart of the summons in Matter No. IRC 4732 of 1997.

In my view, there is a causal connection between the failings identified in the conduct of the defendant and the risk thereby created to the safety of the two operators concerned. Messrs Jovanovic and Husakovic were responsible and sensible workers, as is evidenced by the fact that the defendant considered they should work unsupervised. I am satisfied that had they been adequately trained, instructed and warned to check the pressure gauges and the computer screens to ensure that there was zero pressure within the cooker before attempting to release the lid after a bad cook, they would have done so. It was not sufficient to rely upon any general instruction as to the checking of gauges, (which, in any event, did not include the checking of a trend screen) given years before the event to bring home to the operators the need for rigorous checking of pressures before opening lids, particularly in relation to bad cooks, where an expectation of the absence of pressure may exist. Had they carried out those checks they would have seen from the gauges and the computer screens that the cooker was pressurised. They would not, I believe, have then attempted to release the lid, but would have sought other means of reducing the pressure within the cooker. The accident would not have occurred. I hold, therefore, that the risk to safety was caused by the defendant's breaches of duty I have identified.

It is not for the Court to specify how the defendant's statutory duty could be fulfilled: see Drake at 36-37 and WorkCover Authority of NSW (Inspector Penfold) v Fernz Construction Materials Limited (unreported, Walton J, Vice-President, IRC3235 and 3236 of 1998, 3 May 1999); but better warnings, instructions and training of the operators were required in the instant case. The need to take such steps was, in my view, all the greater where the system of work employed by the defendant was not to supervise the operators on the combination line, but to rely on them to cope with problems and, in particular, to foresee and take precautions against the risks of the job: see Cullen at 218. It was necessary to actively and consistently convey to the operators the need to conduct such checks.

In any event, the mere stipulation of procedures and instructions to carry out the work may not be enough to satisfy the requirements of s15 of the Act. The employer should ensure that procedures and instructions are actively and positively complied with by employees.

In Dawson (at 121 and 125) the failure by a foreperson to supervise a qualified and experienced linesperson to the extent of instructing the employee as to the safest available method (where the employer had laid down a procedure for work and various methods of performing the work were available) would constitute a breach of s15 (although in that matter it was found the failure was that of the foreperson and not the employer).

I note that in an English High Court case, Pope v Gould (HM Inspector of Health and Safety, 20 June 1996, unreported but cited at p44 of Redgrave's Health and Safety, 1998, Third Edition) it was held in England, under the equivalent of s15, that it is not a sufficient defence for an employer to demonstrate that he had instructed the employee to operate machinery in a particular way and the employee had not followed his instructions. There, it was held, that the legislative provision was not concerned solely with the giving of instructions; rather it imposes upon the employer positive duties in relation to ensuring safe systems of work and safe machinery. It was held not to have been sufficient for instructions of a health and safety nature merely to have been given to employees; an employer must also ensure that those instructions are carried out (see also Haynes at 181-182, Kirkby at 53 and McMartin at 8).

Exhibit 11 contains a new procedure entitled "Cooker Lid Removal Procedure - Manual Operation". That document states at paragraph 1.6 that the cooker pressure must be checked both at the cooker pressure gauge and on the computer display "to establish that the cooker is no longer under pressure, i.e. the cooker gauge reads zero (0) and the [computer] display reads zero (0)". The procedure also requires that "at all times during this procedure the cooker shall be considered under pressure until proven otherwise". The document, in the description of the purpose of the procedure, stipulates that the procedure expressly applies to bad cooks. Additionally, a bad cook disposal and conditioning procedures were brought into existence.

The revisions were steps which demonstrably could have been taken in relation to bad cooks before the accident but were not. Instructions in relation to bad cooks could have been specifically conveyed to operators and appropriate training could have been supplied. However, even had that revision taken place prior to the accident and appropriate instructions and training issued, it would have been insufficient, in my view, to fulfil the duty on the defendant which I have identified. In view of the patent danger of opening, or partially opening, a cooker under pressure, as was so dramatically illustrated to Messrs Bannerman, Vanslambrouck and Zolnierczyk, the paramount importance of carrying out the pressure check, identified in the revised manual, should have been brought home to the operators much more forcefully than simply inserting it in a manual or giving an instruction as to the use of such a procedure. The operators were largely self-taught and self-supervised. The experience (or experiences) of Messrs Bannerman, Vanslambrouck and Zolnierczyk was never conveyed to Messrs Husakovic and Jovanovic, although they knew vaguely of an earlier incident. The risk that a cooker charged with product might be under pressure of boiling steam by reason of a phenomenon known to occur from time to time, namely a leaking valve, appears self-evident.

I note, though also not necessary to my decision, that one reason for the defendant's failure to fulfil its duty is that it failed to apply a managerial mind to the assessment of the risks inherent in the job delegated to the two operators. The duty to assess the risk of the task allocated to Mr Husakovic and Mr Jovanovic should have been taken by management and could not be delegated, in particular, to the operators themselves: Cullen at 219. In that regard I note the decision of the English Court of Appeal in R v British Steel Plc [1995] ICR 586 at 592 to 594 where it was held, in relation to parallel UK legislation, that a corporate employer could not avoid liability on the basis that the company at senior management level had taken all reasonable care to discharge its duty by delegating its health and safety obligations.

I should also mention the revision of the manual which took place after, and in the light of, the accident. I note in passing that Mr Jovanovic and Mr Husakovic suggest that one check advised by the revised manual (paragraphs 1.13 to 1.15 "check underfloor bleed valve") would be unsafe. It is unnecessary for me to make findings about this.

I find that the defendant failed to ensure the health and safety of Messrs Husakovic and Jovanovic whilst at work by failing to provide and maintain systems of work that were safe and without risk to health as required by s15 of the Act. They are guilty of an offence under s15 in relation to the charge brought in Matter No. IRC 4732 of 1997, unless able to make out a defence under s53 of the Act. I do not find liability in relation to the charge brought in Matter No. IRC 4733 of 1997 and the defendant is, therefore, not guilty of an offence under s15 of the Act in relation to that matter.

Having determined that the elements of the offence have been established, it is then appropriate to consider whether a defence has been made out under s53 of the Act: ABB Power Transmission Pty Limited v WorkCover Authority of NSW (Insp. Wilson) (unreported, Fisher P, Bauer and Hungerford JJ, CT96/1215, 2 May 1997 at 8).

"It shall be a defence to any proceedings against a person for an offence against this Act or the Regulations for the person to prove that:

(a) it was not reasonably practicable for the person to comply with the provision of this Act or the Regulations the breach of which constituted the offence, or

(b) the commission of the offence was due to causes over which the person had no control and against the happening of which it was impracticable for the person to make provision.

To establish a defence under s53, the defendant must prove, to a civil standard, either that it was not reasonably practicable to comply with the Act or that the commission of the offence was due to causes over which the defendant had no control and against the happening of which it was impracticable to make provision: see Drake at 48-49; Sydney City Council v Coulson (1987) 21 IR 477 at 480; and Italo Australian Construction Pty Limited v Parkes (1988) 24 IR 428 at 431. So much was accepted by the defendant in this case.

It is here that the question of reasonable foresight has relevance. If a defendant is able to demonstrate that the circumstances or causes of the detriment to safety constituting the offence were not reasonably foreseeable, it will generally have thereby established that it was not practicable to take measures to guard against that risk: see Jayne v National Coal Board [1963] 3 All ER 220 at 224 and Shannon v Comalco Aluminium Ltd (1986) 19 IR 358 at 363-364. In WorkCover Authority of NSW (Insp. Mayo-Ramsay) v Maitland City Council (1998) 83 IR 362 at 381, for example, Hill J commented in relation to s53:

If the happening of an event is not reasonably foreseeable it is not practicable to make provision against it. When considering the matter of foreseeability, one should be careful not to substitute reasonable hindsight for reasonable foresight.

This does not remove the obligation, to which I have earlier referred, of an employer to adopt a proactive approach to safety issues which seeks to discover and remedy potential risks to its employees and others. It is not a question of whether the employer did envisage a particular danger, but rather whether it should have.

In this case, the accident was not, as Mr Carson sought to characterise it, a freak occurrence which cannot be explained and which arose in an unknown situation. The risk of valves leaking was both known and foreseeable. The possibility of steam leaking from a defective valve into a container charged with spoiled product and on which the lid was secured so containing the pressure was equally obvious. The possibility of an explosion thereby resulting was also known. These were dangers which the defendant could have, and should have, taken steps to address.

In relation to the reasonable practicability of taking measures to remedy the potential threat to the safety of the defendant's employees, I consider that the formula identified in negligence cases such as Paris v Stephney Borough Council [1950] UKHL 3; [1951] AC 367 at 375-376 is relevant. In short, the issue to be addressed may be expressed as follows: the greater the magnitude of the risk and the greater the gravity of the harm, should the event occur, the higher is the duty to take precautions, even if these are expensive or difficult to adopt. A similar approach was adopted by Asquith LJ in Edwards v National Coal Board [1949] 1 KB 704 at 712:

The onus was on the defendants to establish that it was not reasonably practicable in this case for them to have prevented a breach of s49. The construction placed by Lord Atkin on the words "reasonably practicable" in Coltness Iron Co. v. Sharpe [1938] AC 90 at 94 seems to me, with respect, right. "Reasonably practicable" is a narrower term than "physically possible" and seems to me to imply that a computation must be made by the owner, in which the quantum of risk is placed on one scale and the sacrifice involved in the measures necessary for averting the risk (whether in money, time or trouble) is placed on the other; and that if it be shown that there is a gross disproportion between them - the risk being insignificant in relation to the sacrifice - the defendants discharge the onus on them.

This passage was cited with approval by Watson J in Carrington Slipways Pty Limited v Callaghan (1985) 11 IR 467 at 470 and by Miller CIM in McVernon v Royce Roofing Pty Limited (1988) 25 IR 383 at 385.

In the instant case, I find that the magnitude of the risk was not so small that it could be ignored and that it may be characterised as real or substantial. The gravity of the harm which could have resulted should the event occur was extreme: loss of life was a real and substantial possibility should a lid be loosened or removed on a cooker full of spoiled product and steam at pressure. There is no evidence that the minimum precautions which I have identified of providing training, instruction and warnings to the operators were of prohibitive difficulty or expense. Nor was there evidence that the provision of adequate supervision would create such problems.

Additionally, it follows from the foregoing discussion that, for the purposes of s53(b), the commission of the offence was due to causes over which the defendant had control and against the happening of which it was practical for the defendant to make provision.

Accordingly, the s53 defence fails. I hold that it was reasonably practicable for the defendant to comply with s15.

I find the offence proven in relation to the prosecution brought in Matter No. IRC 4732 of 1997. I find the defendant not guilty of the offence charged in Matter No. IRC 4733 of 1997. I will hear submissions on penalty on a date to be fixed by arrangement with the prosecutor and the defendant.