Industrial Relations Commission of New South Wales in Court Session CITATION : WorkCover Authority of New South Wales (Inspector Rodney Dubois) v Coffey Partners International Pty Ltd. [2004] NSWIRComm 373 PARTIES : WorkCover Authority of New South Wales (Inspector Dubois) v Coffey Partners International Pty Ltd. FILE NUMBER: IRC 5705 of 1999 CORAM: Haylen J CATCHWORDS : Occupational Health and Safety Act 1983 - s 16(1) - defendant providing geotechnical advice for construction of railway link to Sydney airport - construction and design of cofferdam in Cooks River responsibility of other corporations - allegation of failure to assess strength of sandstone to support cofferdam walls - defendant carried out extensive testing and assessments of geotechnical nature - prosecution case expands beyond strict terms of particulars - evidence demonstrates that defendant did make appropriate assessments of strength of sandstone - assessments provided to designer and construction company - expert evidence does not call into question appropriateness of investigation and advice provided by defendant - prosecution case not made out on the evidence - prosecution dismissed HEARING DATES: 05/20/2002; 05/21/2002; 05/22/2002; 05/23/2002; 05/27/2002; 05/28/2002; 05/29/2002; 05/30/2002; 05/31/2002; 02/13/2003; 02/17/2003; 02/18/2003; 02/24/2003; 02/25/2003; 03/04/2003; 03/05/2003; 07/11/2003; 07/23/2003; 11/17/2003; 11/18/2003; 02/11/2004; 02/12/2004; 02/13/2004; 09/28/2004 DATE OF JUDGMENT: 12/06/2004 PROSECUTOR: Mr P Skinner of counsel SOLICITORS: Moray & Agnew LEGAL REPRESENTATIVES: DEFENDANT: Mr R M Goot SC and Mr A Ridley of counsel SOLICITORS: Minter Ellison JUDGMENT: - 91 - INDUSTRIAL RELATIONS COMMISSION OF NEW SOUTH WALES IN COURT SESSION CORAM: HAYLEN J 6 December 2004 Matter No. IRC 5705 of 1999 WorkCover Authority of New South Wales (Inspector Rodney Dubois) v Coffey Partners International Pty Ltd. Prosecution under s 16(1) of the Occupational Health and Safety Act 1983. JUDGMENT [2004] NSWIRComm 373 BACKGROUND 1 During 1997, the rail link between Sydney's train service and the airport was well under way. A particular part of that rail link was known as the Cahill Park, Cooks River Road Railway Tunnel, located at Arncliffe. The tunnel being constructed was to accommodate two railway lines and the tunnel was to cross under the Cooks River. The method of construction was described as a cut and fill approach whereby a very large trench was dug below the level of the river with box sections ultimately inserted, thus making the tunnel. The large trench was then to be backfilled but at Cahill Park, because the tunnel had to cross under the river, the construction of a cofferdam was required. This cofferdam was constructed by driving steel piles into the sandstone of the riverbed and then pumping out water to allow excavation to take place. 2 This part of the underground tunnel in the airport rail link was being constructed by a joint venture between Transfield Australia Pty Ltd and a French company named Bouygues: the joint venture was referred to as TBJV. Transfield Australia Pty Ltd had the contract to construct the underground railway through one of its divisions, Transfield Constructions Pty Ltd. In relation to the construction of the cofferdam, Transfield Constructions Pty Ltd contracted to other companies: the designer of the cofferdam was Connell Wagner Pty Ltd, which in turn was advised by the defendant in these proceedings, Coffey Partners International Pty Ltd. Coffey Partners advised on geotechnical matters including the composition of bedrock into which the interlocking steel piles were to be driven. Coffey Partners was well known as an expert in this particular field. 3 The cofferdam at Cahill Park had been constructed for some months by October 1997 and excavation was well under way. Coffey Partners had been engaged from 1995 to report to the Joint Venture on site conditions, and in 1996 to give advice to Transfield about geotechnical aspects of the project. In 1997, Coffey Partners was providing geo-technical data and advice on and off site mainly through Dr Peter Redman. This advice was being supplied to Connell Wagner. 4 From as early as April 1997, issues arose at the construction site in relation to the structure of the cofferdam. By 17 October 1997, a substantial amount of excavation had taken place up to the eastern wall of the cofferdam. On 18 October 1999, there were leaks into the cofferdam and shotcrete within the dam was breaking away. These problems continued through 19 and 20 October 1997 and remedial measures were taken on site. A number of employees working on the site expressed their concern and meetings were held to discuss those issues: at one of those meetings Dr Redman attended and spoke to the men about the safety of the site. On 22 October 1997, at approximately lunchtime, the eastern wall of the cofferdam failed and an inrush of water flooded the tunnel construction site. A number of employees had already left the bottom of the cofferdam for lunch but two employees, Darryl Grey and Juan Bahamondz were swept through the tunnel area by the force of the water and were seriously injured. 5 In October 1999, Inspector Dubois of the WorkCover Authority commenced proceedings against the defendant Coffey Partners International Pty Ltd claiming a breach of s 16(1) of the Occupational Health and Safety Act 1983 arising out of the collapse of the eastern wall of the cofferdam on 22 October 1997. The Inspector alleged that, between 21 and 22 October 1997 at the Cahill Park Cooks River railway tunnel construction project at Arncliffe, Coffey Partners International Pty Ltd, the defendant, had failed to ensure that persons not in its employment and, in particular, Darryl Grey and Juan Bahamondz, were not exposed to risk to their health and safety arising from the conduct of its undertaking, namely, the provision of geotechnical advice for the design of the cofferdam while they were at the defendant's place of work. The particulars of the charge were: (a) the defendant at all material times employed personnel at the construction site; (b) the defendant was engaged by Connell Wagner Pty Ltd to provide geotechnical advice as to the strength and integrity of the sandstone bedrock at the base of the cofferdam wall at the construction site; (c) the defendant failed to adequately assess the strength and integrity of the sandstone bedrock at the base of the cofferdam wall; (d) the defendant failed to assess the capacity of the sandstone bedrock to provide support at the toe of the cofferdam wall; (e) as a result of these failures, Darryl Grey and Juan Bahamondz suffered multiple injuries; (f) and, as a result of the failures, a number of workmen were placed at risk of injury. 6 The defendant has entered a plea of not guilty to this charge. The hearing of the evidence and submissions relating to this alleged breach has occupied some 24 hearing days. The pages that follow deal with the essential elements required to be made out by the prosecutor in order for a finding of guilt to be made against the defendant. In view of the number of people involved and the technical industry terms used, a glossary appears at the conclusion of the judgment. 7 Very early in the hearing the defendant made the following admissions: (a) at all material times, the prosecutor was a duly appointed inspector for the purposes of the Occupational Health and Safety Act 1983; (b) at all material times, the defendant was duly incorporated; (c) at all material times, the new southern railway tunnel was being constructed through Cahill Park and that the principal contractor of the project was the TBJV (Joint Venture) which had contracted Transfield Constructions Pty Ltd to construct that section of the tunnel; (d) Transfield Constructions Pty Ltd had contracted Connell Wagner to undertake the design of the sheet pile cofferdam; (e) the defendant was engaged by Connell Wagner to provide geo-technical advice in relation to the project; (f) at all material times, the defendant employed personnel providing certain geotechnical advice in connection with the cofferdam design; (g) at all material times the defendant was an employer, that the site of the cofferdam was the defendant's place of work as at 21 October and 22 October 1997, and on those days the cofferdam was part of its undertaking. PROSECUTION CASE 8 The prosecutor accepted that it was first necessary to consider whether the elements of the alleged offence had been established beyond reasonable doubt before considering the defence case. 9 In the present prosecution, the essential particulars were found in paragraphs (c) and (d), namely, a failure by the defendant to adequately assess the strength and integrity of the sandstone bedrock at the base of the cofferdam wall, and the defendant's failure to assess the capacity of the sandstone bedrock to provide support at the toe of the cofferdam wall. The defendant, as a specialist in the field, had been giving geo-technical advice to Connell Wagner and Transfield Constructions Pty Ltd relating to the design of a cofferdam. There was no precise evidence of the terms of engagement of the defendant by Connell Wagner but the evidence supported a conclusion that the defendant was engaged on an ad hoc basis to provide advice to both Connell Wagner and Transfield Constructions Pty Ltd as the need arose and at the critical time was providing such advice on a daily basis. This advice related to the state of the rock into which the cofferdam piling had been driven and the integrity of the support provided by that rock. The prosecutor accepted that the defendant did not design the cofferdam nor was it engaged to do so, but it was involved in the design of the cofferdam by providing geo-technical advice on which the designers relied: Connell Wagner was responsible for the design of the cofferdam and Transfield Constructions Pty Ltd was responsible for its construction. In addition, the design of the cofferdam was said to include "continuing to re-visit" as rock was taken away, and checking the status of the cofferdam as the physical circumstances changed. 10 The evidence of a Transfield Constructions Pty Ltd employee, Mr Shepherd, was relied upon by the prosecutor as establishing that there was never a fixed design for the cofferdam and in particular that was so in April 1997. The process adopted was "design and construct" and to develop the design as construction proceeded. Thus, during the construction phase, there were a number of proposals considered including the installation of rock bolts and discussion as to the dimensions of the berm to be left at the foot of the sheet piles at the eastern wall. These were suggestions made prior to being able to inspect the toe of the wall. Mr Shepherd was the project manager for Transfield Constructions Pty Ltd and was engaged in that capacity at the Cahill Park site. He described a "design and construct" project as one where the client gave conceptual design and design parameters to the contractor to consider them and to then design and construct the structure including the detailed design, gathering all the design inputs and producing a set of design outputs or deliverables which were handed to the client for review. At Cahill Park, it was a typical "design and construct" contract where the design was occurring throughout the course of the works should circumstances required it. There was an observation method introduced and that method was referred to by both Connell Wagner and the defendant: it operated on the basis that, once the site had progressed to certain levels in the excavation, test pits would be excavated and the geological model or assumptions would be verified. The geological model was based on interpretation of the ground between boreholes - it was referred to during the driving of piles to obtain a verification of rock levels. The geological model was developed by Coffey Partners. 11 Mr Shepherd agreed that the role of Connell Wagner in relation to the cofferdam project was to certify construction in accordance with its design. The method of supporting the cofferdam at the toe of the sheet piles was the responsibility of Connell Wagner, according to Mr Shepherd. There had been discussions, including those on 30 September 1997, about the use of rock bolting. The defendant in discussions with Connell Wagner had come up with a design of leaving a three metre rock bench in front of the walls in lieu of rock bolts. That proposal was advised to Transfield Constructions Pty Ltd by Connell Wagner, who had also provided drawings from time to time regarding the method of fixing the toe of the wall - these drawings were provided to Transfield Constructions Pty Ltd. 12 In September 1997, although there had been a proposal to rock bolt the eastern wall, there were difficulties in doing so and, in discussions between Transfield Constructions Pty Ltd, Connell Wagner and the defendant, it was decided to leave a much wider berm and to re-visit the matter as the excavation proceeded. 13 By 21 October 1997, Dr Redman had advised Connell Wagner that rock bolts were to be used to secure the eastern wall. In his evidence, Dr Redman stated that after reviewing the state of the excavation on 21 October 1997, it was decided to bypass the inspection stage and go straight to the use of rock bolts. If the inspection stage had been maintained, a test pit would have been sunk in order to decide whether or not the strength of the sandstone was sufficient to support the wall and whether rock bolting was necessary. 14 The prosecutor submitted that the defendant may have been open to a charge dating from 18 October 1997, because of all the warning signals being given: shotcrete was falling from some of the walls of the cofferdam (although not the eastern wall), there were leaks in other walls and a leak on 21 October 1997 at the eastern wall; there were negative readings on the load cells and, over the weekend, struts (or whalers) had been inserted to strengthen the cofferdam. Given those warning signals, the defendant had failed to investigate and adequately assess the strength and integrity of the sandstone bedrock at the base of the cofferdam. This breach certainly occurred on 21 and 22 October 1997, as demonstrated by the failure of the eastern wall and the inrush of water, which led the site to be totally flooded for a number of months. That was an obvious risk to the persons employed at the site. 15 Asked to indicate how this breach crystallised on 21 and 22 October 1997 the prosecutor submitted that it was the defendant's failure to assess adequately and investigate the quality of the sandstone and to inspect the width of the berm at the base of the eastern wall. In August or September 1997, Dr Redman for the defendant had expressed concern about over-excavation of the berm and, further, in September 1997, there had been a decision not to rock bolt at that stage but to maintain a berm of 3 to 3.5 metres. The prosecutor pointed out that the failure of the defendant, through Dr Redman, was to inspect the berm on 21 and 22 October 1997 in circumstances where the evidence was that the berm had been trimmed and was between 2 and 2.5 metres rather than 3 to 3.5 metres. 16 The prosecutor also identified another way in which the defendant had committed the alleged breach. Dr Redman had addressed a meeting of workers on the site at approximately lunchtime on 21 October 1997. It was submitted that the evidence supported a finding that Dr Redman had allayed the fears of the workforce about the safety of the site: however, the site was unsafe because of the falling shotcrete, the water leaks and the need to use of the struts. Indeed, Dr Redman had assured the workers that the entire site was safe. 17 Although it was accepted that, on 21 October 1997, Dr Redman had decided to bypass the taking of core samples and the inspection stage and go directly to rock bolting the eastern wall and that such steps that may amount to assessing the adequacy of the rock to support the wall, it was submitted that another aspect of the alleged breach was Dr Redman's failure to assess the urgency of the need for rock bolting the eastern wall. It was to be remembered that the wall failed within 24 hours of this recommendation for rock bolting, and that was evidence of the risk. As events showed this remedy of rock bolting for that risk was completely inadequate. An adequate assessment of the situation would have required a recommendation that there be immediate rock bolting and advice that the site was to be cleared of all workers while that took place. In this respect, he had failed to assess the impending failure of the dam. It was put by the prosecutor that Dr Redman, as an expert in the field, should have been in a position to know, especially from the surrounding circumstances, that the cofferdam was in danger of immediate failure. From mid-September 1997, he had expressed concern about over excavation of the berm and there were events from 18 October 1997 of shotcrete falling and leaks occurring, which had caused disquiet amongst the workforce. Dr Redman was at the site on Saturday 18 October 1997 and had an opportunity on that date to inspect the berm. He had passed by it and had looked at the eastern wall without a close inspection. His evidence was that there was an amount of soil and debris on the berm and that it was being cleared. 18 It was accepted by the prosecutor that the recommendation made in a written note from Dr Redman to Connell Wagner on 21 October 1997 made it clear that the toe support at the base of the eastern wall could not be relied upon until anchor bolts were installed and that support was to be given before further excavation or removal of material. The inadequacy of this advice nevertheless, was its lack of urgency. Dr Redman apparently was simply unaware of the imminent risk of the failure of the eastern wall. In the prosecutor's submission, the nature of the advice being provided by the defendant and the professional expertise and duty of Dr Redman was such that he should have been aware that the eastern wall was in danger of failing catastrophically and immediately. It was conceded that there was no expert evidence to that effect but it arose from the combination of circumstances on the site. Those circumstances were again identified as the movement of the wall, the occurrence of leakages, the fact that there was a possibility that the berm width had not been left as intended, the possibility of fracturing from the driving of the piles into the sandstone and the fact that Dr Redman should have known that there was an inherent weakness of the sandstone, that it was essentially category HAW3, not HAW2 as he seems to have thought. It was very weak almost crumbly sandstone. Further, Dr Redman should have paid more attention to the concerns being expressed by this experienced workforce before assuring them that the site was safe, and doing so without making a proper assessment of the sandstone at the base of the eastern wall and by checking the width of the berm at the base of the eastern wall. The risk of collapse had arisen by the time the excavation had reached the final geometry of the berm and that occurred on 21 October 1997. It was stated by the prosecutor that the characterisation of the sandstone as HAW3 (and weaker than HAW2 as assessed by Dr Redman) was to be gleaned from the expert evidence of Professor Carter, although the Professor did not use those terms. There was no evidence, it was conceded, as to what different design or approach to construction would have been taken had it been known to Connell Wagner and Transfield Constructions Pty Ltd that, in fact, the sandstone was the weaker HAW3, rather than the HAW2 assessed by Dr Redman. The prosecutor frankly conceded that the significance of the sandstone being HAW3 rather than HAW2 could not be established because it was not addressed by any expert evidence. It was submitted that, in any event, that gap in the evidence had no significance for the prosecution, as such evidence would be relevant to the design or design changes but not to the nature of the breach as alleged in the charge. 19 The prosecutor's case essentially was that the sandstone was weak and Dr Redman should have known that fact. It may well have been that, having established the weakness of the sandstone, the defendant could then have addressed ways in which to meet the risk that was posed to those working below sea level in the cofferdam. The defendant's obligations under the Act were not met nor was his professional obligation met once Dr Redman made the recommendation on 21 October 1997 that there was to be no further excavation or removal of material and that toe support could not generally be relied upon until anchor bolts were in place. 20 In further explanation of the "design and construct" approach, the prosecutor submitted that the evidence established that different solutions were being looked at and experimented with from time to time so that, if one method did not work, another method was considered. This was the effect of the evidence of Mr Shepherd from Transfield Constructions Pty Ltd and, in particular, was demonstrated by rock bolting in the corner of the cofferdam which was not pursued because it met with an inrush of water, and that was when the defendant advised that a three metre berm should be left at the base of the wall. The test pits which had been dug by the defendant were used to assess the integrity of the rock. The defendant's approach was, rather than bolting the cofferdam immediately, excavation could continue leaving a 3 to 3.5 metre berm and, as excavation got closer to the final level, there would be inspections to ascertain whether there was sufficient rock to hold the toe of the wall in place and, if insufficient rock support was available, then rock bolting would take place. The prosecutor noted that this was the evidence of Dr Redman. The prosecutor conceded that there was no evidence that the test pits dug by the defendant were in such a situation as to expose to the defendant the weakness of the sandstone at the eastern wall. The test pits were dug in different areas and some areas showed different quality rock. It seemed an open inference from the evidence that the sandstone at the collapsed wall was weaker than the sandstone found in the test pits. The prosecution case asserted that Dr Redman should have known that the sandstone at the eastern wall was very weak. 21 Mr Shepherd's evidence was that the defendant had provided Transfield Constructions Pty Ltd with ongoing geotechnical advice during the construction of the cofferdam and that the method of toe fixity employed at the cofferdam was determined by Connell Wagner and the defendant. Mr Shepherd stated that Transfield Constructions Pty Ltd ensured that the sheet pile achieved the design embedment depth by being supplied with the piling records which were compared to the driving criteria as specified by Transfield Constructions Pty Ltd consultants. This evidence was relied upon by the prosecutor as deeply involving the defendant in the design aspects of the cofferdam. Transfield Constructions Pty Ltd had asked Dr Redman to check how the piles had been driven into the rock to see if it was consistent with his estimate of the hardness of the rock. It was not asserted by the prosecutor that the defendant was laying down precisely how far the sheet piling was to be embedded in the rock but that the defendant was considering the pile driving records and was having input with Connell Wagner and Transfield Constructions Pty Ltd in relation to that matter. The reviewing of those records could not be separated from a consideration of the adequacy of the sandstone bedrock. 22 Another aspect of the evidence which demonstrated that Dr Redman should have been aware that the eastern wall was in imminent danger of failing was the presence of water leaking through the eastern wall on 18 and 19 October 1997. Mr Shepherd said that he raised the presence of a leak in the eastern wall with Dr Redman on 18 October 1997. It was also said that on 19 October 1997 the leak had become so bad that the excavator operator had placed the excavation bucket in front of the wall where the leak was present. However, the excavator operator denied doing so. The presence of leaks at the eastern wall prior to 21 and 22 October 1997 and the failure of Dr Redman to inspect the width and adequacy of the berm in front of the eastern wall were at the heart of the breach alleged. The fact that the defendant skipped over the usual step of conducting an inspection of the rock using test pits but went straight to a recommendation of rock bolting without any further excavation must have meant that the defendant held concerns about the adequacy of the rock to hold the eastern wall and the possibility that it may fail. While it might be conceded that this was a prudent step, it was one taken in the context of the job continuing. On this approach, it was accepted by the prosecutor that the fact that there was a leak in the cofferdam did not necessarily mean that the dam was in danger of imminent failure but it was a factor to be taken into account by the defendant. There would not necessarily be a breach by the defendant every time there was a leak in the dam and where the defendant had not taken steps to clear the workforce from the site: it was a question of degree in each case. 23 The defendant, through Dr Redman, had not told the men at the lunchtime meeting on 21 October 1997 that the defendant had advised that support for the wall could not be generally relied upon until rock bolting had been completed or that there was to be no further excavation. Earlier, Dr Redman had complained of over-excavation and he was therefore aware that Transfield Constructions Pty Ltd workers were less than scrupulous in adhering to excavation instructions. Further, on Saturday 18 October 1997 there was a conversation between Mr Shepherd, Mr Di Cesare and Dr Redman where Mr Shepherd asked whether there was a risk of catastrophic failure to which Dr Redman replied that, if the north wall was alright, the site should be alright. Messrs Shepherd and Di Cesare did not recall any such qualification being made by Dr Redman at the time. While this conversation did not occur in the days covered by the charge, it was a serious concern expressed some days earlier to Dr Redman asking if the circumstances were such that there could be a catastrophic failure. That conversation formed part of the background which required Dr Redman to make a proper assessment of the strength of the rock supporting the cofferdam and an inspection of the berm in front of the eastern wall. The defendant and Connell Wagner had been called in on Saturday 18 October 1997to advise on remedial measures at the site. The advice given was to install a strut and Dr Redman returned early the next day to inspect that work. 24 On 21 October 1997, there was a meeting with the men: Dr Redman, while asserting that he was primarily dealing with the reasons the shotcrete was falling and the remedial measures taken in relation to that failing, did refer to the fact that there would be toe anchors on the eastern wall. The prosecutor submitted that it was abundantly clear that Dr Redman not only addressed the men as to the safety of the southern and northern walls but also addressed the safety of the eastern wall. The evidence supported a finding that he gave an assurance to the men that the whole of the site, including the eastern wall, was safe. 25 The process of monitoring the excavation at the cofferdam was further highlighted by a Connell Wagner report to Transfield Constructions Pty Ltd dated 30 June 1997. That report noted that two issues critical to toe support of the pile were the quality of the HAW2 rock and the existence of large blocks or wedges of jointed sandstone. It was noted that the failure of wedges was likely to be less of a problem given the small excavated depths beneath the bench level. Connell Wagner stated in this report that the methodology for determining and addressing toe support was essentially the same with the exception that a row of rock anchors could be required. The report pointed out that lack of toe fixity could occur if the pile was driven into very jointed HAW2 or if the driving crushed the HAW2. It was known that the pile was embedded the required distance into HAW2 for toe fixity but, if the quality of the HAW2 was poor, it was unlikely that any effect would be noticed on the pile fixity until the rock excavation began near the pile toe. Although it was thought to be unlikely to be necessary, a methodology was proposed whereby the lack of pile fixity could be monitored as excavation occurred down to the rock and lack of pile fixity flagged from observing the load/deflexion characteristics. The report then proposed that, throughout the excavation of the cofferdam, there would be monitoring of the pile movements and strut forces by way of compression load cells. The critical level was around 94.5 metres and load/deflexion gradients could be continually monitored for any sudden change. The report considered the position should excavation proceed towards the base without any sudden change in load/deflexion, then potential jointing or poor quality rock would need to be identified and appropriate remedial action taken. Connell Wagner and/or Coffey Partners were to review the depth and quality of the rock to determine if toe support was required. If toe support was required then rock bolting would be used. The bench during excavation would be regularly inspected by Connell Wagner/Coffey Partners to determine the need for additional rock bolting for support against wedge failure. 26 This report was consistent with Dr Redman's evidence as to what he did at the site and the methods adopted, except that the defendant "skipped" the inspection stage and went straight to rock bolting as a recommendation on 21 October 1997. The significance of the concluding part of the June report was that, during September 1997 and by October 1997, the methodology had completely changed - the focus was upon maintaining a conservative width of berm in front of the eastern wall. 27 The prosecutor accepted that there was no expert evidence to the effect that Dr Redman should have demanded that Connell Wagner provide the design parameters to the defendant to enable the defendant to assess the adequacy of the design. The prosecution was not that type of case. Dr Redman was not being asked to advise whether a cofferdam built to Connell Wagner's design would collapse. It was not the prosecutor's case that Dr Redman was slapdash or cavalier, but that he missed a fatal step or several fatal steps. He did not factor in the possibility of disobedience or mistakes by Transfield Constructions Pty Ltd or the workers who excavated the berm. He did not factor in the possibility that the designer might have got something wrong or would design something wrong. He did not ask either Transfield Constructions Pty Ltd or Connell Wagner about these matters. It was a fatal flaw in the defendant's approach that it was unaware of the weakness of the sandstone and its lack of support for the eastern wall. The prosecutor also accepted that this was not a case where an expert had been called to state a finding that very weak sandstone should have been discovered by the defendant and that is the reason the eastern wall of the cofferdam failed. It was frankly accepted that the prosecution was unable to establish the precise reason for the eastern wall collapsing. 28 In October 1997, the defendant had set out the places where test pits would be dug and proposed that those test pits would be angled. This approach was abandoned by the defendant in favour of going straight to the rock bolting option. The prosecutor submitted that, if this usual step had been taken, it may have and was likely to reveal that the sandstone around the sheet piling which failed was extremely soft. Inspector Vierow had inspected this sandstone after the incident by diving to the area: he was able to penetrate the very weak sandstone with a knife. Dr Redman had bypassed the inspection stage and, without any knowledge of the strength of the sandstone at the eastern wall, had decided that rock bolting was appropriate because of what he had observed by way of leakage and slipping shotcrete on the northern and southern walls of the cofferdam. He had discovered nothing about the clear weakness of the sandstone at the eastern wall. 29 Dr Redman had been cross-examined on the report prepared by Professor Carter. Using the findings made by Professor Carter, Dr Redman stated that the samples referred to in the report placed the sandstone at the eastern wall at the lower end of the HAW2 range. It was therefore not HAW3, and the prosecution could not assert on the evidence that the sandstone at the eastern wall was HAW3 but it did persist with the submission that it was very weak sandstone. 30 On Monday 20 October 1997, Dr Redman had not been on site but another member of the defendant firm had attended. Dr Redman made a file note on that day based on his own observations and what may have been told to him by Mr Parker of Coffey Partners. In that note, Dr Redman spoke about reviewing conditions on the eastern face saying, "Need for toe support and/or control of rock at toe". Dr Redman accepted that this was a live issue in October 1997 because the defendant had yet to carry out an inspection of the eastern area of Cahill Park including the eastern wall and there was still a need to consider the conditions that might be there. Dr Redman said that the review was indirect in the sense that, from what he could see on the northern and southern walls, had led him to a conclusion as to what he would recommend for the eastern face. He could not see any purpose in waiting for the inspection stage to be completed when what he saw elsewhere on the site suggested that rock bolting would be required. In the submission of the prosecutor, at this point there was a clear breach by the defendant because, having acknowledged that there was a live issue to review the eastern wall the decision to rock bolt was taken without looking at the eastern wall but rather by considering what had been seen on the southern and northern walls. All that Dr Redman had to do was push aside some of the spoil on the berm at the eastern wall to discover that it was not 3, 3.5 or 4 metres but had probably been trimmed to as low as 2 or 2.5 metres. If that inspection had occurred and that state of the berm discovered then rock bolting alone was not sufficient: the site would have had to be evacuated while the rock bolting took place. If Dr Redman had taken the time to inspect the eastern wall and had found only a 2 metre berm he would then have known that the wall was ready to collapse. 31 On Saturday, 18 October 1997, Dr Redman had been brought to the site to advise Transfield Constructions Pty Ltd. Connell Wagner had also been brought to the site by Transfield Constructions Pty Ltd. Dr Redman's evidence was that Mr Shepherd on behalf of Transfield Constructions Pty Ltd had drawn attention to shotcrete falling from the northern and southern walls and had asked if the cofferdam was in danger of collapsing. Dr Redman asked whether there had been any movement in those walls and being assured there had not been then stated that he was not concerned about the dam failing. In the course of this discussion, he was informed that there was water coming in from the eastern wall but that the grouting company, Ground Consolidated Pty Ltd, was attending to that problem. He walked by the eastern wall and saw that grouting occurring. The prosecutor accepted that there was probably no evidence that, on 21 October 1997, Dr Redman had been informed that the water was "surcharging" at the eastern wall. In the prosecutor's submission, knowledge of the leak on Saturday night 18 October 1997, together with the collapsing shotcrete and the need to install struts were all necessary parts of the background and knowledge of Dr Redman and, if he had also looked at the inadequate size of the berm on 20 or 21 October 1997, he would have realised that the eastern wall was in imminent danger of failure. During addresses, the prosecutor accepted that the defendant's assessment that the sandstone was HAW2 was one assumed from a number of factors and could not be guaranteed until the excavation stage reached the toe of the cofferdam walls. The prosecutor's submission was that this assumption or presumption was relied upon by the defendant but there was no re-assessment, especially as the excavation got closer to the toe of the cofferdam. Simply concluding that rock bolts should be installed by looking at everything other than the eastern wall was not really a re-assessment. THE EVIDENCE 32 Mr Shepherd was the project manager with Transfield Constructions Pty Ltd and was involved with the cofferdam project. He had informed Inspector Dubois that Transfield Constructions Pty Ltd had been supplied with ongoing geotechnical advice during the construction of the cofferdam by the defendant. He said that representatives of the defendant attended at the site in relation to problems with all water ingress at the cofferdam. Connell Wagner and the defendant had advised Transfield Constructions Pty Ltd on the method of toe fixity and Transfield Constructions Pty Ltd had relied on their expert advice. 33 On 18 October 1997, Mr Shepherd said that representatives of Connell Wagner and the defendant were called to the site because there was a "split clutch" and water ingress in the northern wall. That was injected with polyurethane and plated. Connell Wagner and the defendant also directed the installation of additional temporary strutting between the northern and southern walls. There was also a "minor leak" at the eastern wall. 34 During the interview with the Inspector, Mr Shepherd said that there was an active safety committee on site and that Transfield Constructions Pty Ltd employees had raised concerns regarding the northern and southern wall of the cofferdam and that Transfield Constructions Pty Ltd had then sought advice from their experts. Transfield Constructions Pty Ltd had been told that there was not a risk of catastrophic failure of the cofferdam. That assurance had been given by Dr Redman on behalf of the defendant. The same question had been asked of the Connell Wagner representative but no answer had been given. 35 In oral evidence, Mr Shepherd said that he had informed Dr Redman that there was water bubbling through the eastern wall when they spoke on Saturday 18 October 1997. The water was bringing a sandy deposit through the walls. Dr Redman had said that it was more a construction nuisance issue and it could be dealt with by means of pumping. In earlier written answers to questions raised by WorkCover, Mr Shepherd denied that there was excavation of the rock in excess of the defendant's recommendation or in the vicinity of the eastern wall, although he accepted that there were occasions when excavation was carried out in excess of the defendant's recommendations. 36 It was understood by Transfield Constructions Pty Ltd that Connell Wagner had a "no toe fixity model" as part of their design. It was also understood that Connell Wagner had proposed to monitor the forces on the sheet piles of the cofferdam by use of compression load cells. Although Mr Shepherd did not feel qualified to answer questions about the design of the cofferdam, he understood that there was an assumption that there was no toe fixity from the rock and that other options might have to be developed. The installation of the load cells was arranged by Transfield Constructions Pty Ltd and the monitored results were supplied to the Joint Venture and to Connell Wagner but not to the defendant. 37 Mr Shepherd understood from discussions with Dr Redman that a three to four metre berm was to be left, including around the eastern wall. In mid-August 1997, Dr Redman, in a conversation with Mr Shepherd and Mr Di Cesare spoke about the fact that the berm had been over-excavated under the Princess Highway. Dr Redman emphasised that there had to be a three to four metre berm. He advised that there had to be immediate backfilling and rock bolting at that point: that advice was acted on. 38 On 1 October 1997, Transfield Constructions Pty Ltd received an instruction from Connell Wagner concerning the geometry of the berm in light of the information from the test pits. Connell Wagner advised that a berm of 3.5 metres be left in front of each sheet pile. Mr Shepherd understood that there would be an inspection of the face to establish whether there were any faults or jointing in the rock and at that stage consideration would be given to rock bolting the bench. Mr Shepherd understood from discussions that the berm would then be cut back to three metres and then back to two metres. Mr Shepherd was unaware of any inspection of the berm of the eastern wall prior to him leaving the site on 19 October 1997. 39 On 18 October 1997, Dr Redman was present at the site and had a discussion with Mr Shepherd about the falling shotcrete. Mr Shepherd accepted that Dr Redman had asked him what was happening at the eastern wall and that Mr Shepherd replied that water had been coming in during the day and that a company called Ground Consolidation were working on it. Mr Shepherd said there were two conversations about water at the eastern wall and that was one of them. Mr Shepherd also accepted that, when he asked Dr Redman whether the dam would come crashing down around his ears or that there could be a catastrophic failure he was referring to the northern and southern walls of the cofferdam. The answer to the question was Dr Redman's advice that it would be prudent to install struts. Mr Shepherd's recollection was that up to 18 October 1997 there had been no excavation of the berm at the eastern wall. On 18 October 1997, Mr Shepherd said that it would have been very difficult to see the berm in front of the eastern wall. 40 Dr Gomes was a consulting civil engineer employed by Connell Wagner and was involved in the Cahill Park project, including the construction of the cofferdam. He said that Connell Wagner had been engaged by Transfield Constructions Pty Ltd as its design engineer for the project. The parameters for the Cahill Park cofferdam had been provided by the defendant Coffey Partners, who were the geotechnical engineers for the project. While Connell Wagner was qualified to design the cofferdam it relied upon experts such as the defendant in relation to geotechnical matters. Dr Gomes stated that he understood that the defendant was well regarded and was an expert in the field of geotechnical matters, and that Dr Redman was, in particular, known as an expert in the field of driving sheet piling in Sydney sandstone. 41 Dr Gomes had not visited the site prior to the accident in mid-October 1997 but was aware of the arrangements whereby Connell Wagner were to design and provide advice to Transfield Constructions Pty Ltd. 42 Dr Gomes agreed that part of Connell Wagner's commission was, where necessary, to design toe anchors for the sheet piles and to assess the need for rock bolting or putting in toe anchors in the sheet piles. The sheet piles and box piles had been driven through water and sub-surface of water into the ground surface which the defendant had indicated to Connell Wagner was Hawkesbury sandstone. A pile driver was used to hammer the piles at a low rate which was determined by having regard to the sub-surface they were to penetrate. Transfield Constructions Pty Ltd undertook the pile driving. The function of the cofferdam was to create a dry working area. Connell Wagner had advised that a berm of 3.5 metres be kept as it was regarded as providing sufficient toe support to enable an inspection to be made before any further excavation occurred. This approach allowed an inspection, as the excavation proceeded, to determine the quality of the sandstone. Depending on the quality of the sandstone, a decision was then to be made as to the use of rock bolts. 43 Dr Gomes was familiar with the classification of Hawkesbury sandstone and accepted that sandstone was not homogenous and that within classes III and 1V there was a range of rock strength. The differences in Class III and Class IV sandstone made it necessary for the inspection process to be undertaken and for a better assessment to be made of the quality of sandstone existing at the inspection site. Early in the project, the defendant had conducted an extensive investigation of the rock through which the project passed and boreholes had been dug as part of that exercise. There was at least one borehole which was in close proximity to the eastern wall of the cofferdam: a number of other tests, including seismic tests, were undertaken by the defendant. 44 Dr Redman had recommended the method of toe support, which was penetration of the sheet piles 300mm into HAW2 sandstone. That advice from the defendant had been used in the design of the cofferdam. Dr Gomes understood HAW2 referred to moderately weathered sandstone and that HAW3, which was also known as Class V sandstone, was very weak sandstone, "almost like soil". Dr Gomes understood that HAW2 material was quite variable in nature. 45 Dr Gomes was not aware of all of the documents that may or may not have been supplied to the defendant during the course of the project. He was unaware whether the load cell readings had or had not been passed on to the defendant. 46 In relation to remedial action, Dr Gomes said that safety considerations were clearly paramount and that the "observational technique" employed in this design required fairly prompt action in the event of circumstances arising such as the requirement for rock bolting. Dr Gomes said that the classification of sandstone was well understood and that the classification used at this site would have HAW1 as the strongest sandstone, HAW2 the intermediate strength sandstone and HAW3 as the weakest material. 47 Professor Carter was Challis Professor of Civil Engineering at Sydney University. His speciality was geotechnical engineering and he had prepared a report on the testing of sandstone at the Cahill Park site. An investigation was conducted to examine the ground conditions at the base of the failed sheet piles at the eastern end of the cofferdam and samples were recovered for laboratory testing and to enable measurements to be made of the strength of those samples in laboratory conditions. 48 During the investigation, five core holes were drilled into the sandstone where the sheet piling had been driven at the eastern wall. Observations of the exposed faces of the rock and the recovered core revealed that the upper 300 - 400mm was extremely weathered sandstone grading into highly weathered sandstone at about the level of the toe of the box pile. Visually, the sandstone was assessed as having very low to extremely low strength. It contained weaker thin zones or partings that appeared to be composed predominantly of sand clay. The presence of these weaker zones made the recovery of intact samples difficult as the core tended to separate at the weak zones. Inspections of the hole formed by the drilling revealed the presence of cracks in the rock mass in the toe region of box pile 26. The cracks appeared to be relatively recent, possibly resulting from the driving of the steel piling. Visual inspection of the boreholes and the core recovered from them did not reveal the presence of natural rock joints either at the level of the toe of the box pile or at any other level. The Professor indicated that, during laboratory testing, obtaining suitable specimens was difficult because of the tendency of the samples to disintegrate during preparation. 49 Professor Carter explained the nature of the tests conducted. From three identified samples, the Professor calculated unconfined compressive strength at 2.3, 2.5 and 3.3 MPa. 50 Mr Richard Clark was a professional engineer with long experience who had provided a report to WorkCover on the collapse of the cofferdam. His report was subjected to close analysis and criticism by counsel for the defendant and, in particular, on his limited capacity to make any assessment as to the strength of the sandstone in the area where the cofferdam wall collapsed. His evidence was allowed on a limited basis on the understanding that he did not suggest that he had geotechnical expertise. His qualifications and experience did qualify him to conclude that the sandstone was quite weak in the area of the cofferdam wall that failed. Mr Clark's conclusion was that the cofferdam wall collapsed because the material into which the toe of the piles was driven did not have sufficient strength to resist the horizontal forces applied by the toe of the piles. 51 Mr Andrew Shirley was a consulting engineer operating under the name Shirley Consulting Engineers Pty Ltd. Both Mr Shirley and his company had been involved for many years in geotechnical investigations and the design and supervision of engineering works related to industrial and domestic construction. He had been engaged by the WorkCover Authority of New South Wales to conduct an investigation into the collapse of the cofferdam at the Cahill Park site. As a result of that engagement, he had produced a report which was supported by a number of documents and photographs. 52 Mr Shirley's report noted, as an overview, that the bedrock underlying the over-burdened soil material in the vicinity of the excavation varied significantly in quality and type, with a number of clay seams, rock joints, bedding planes and similar rock mass discontinuities being present. The nature of the bedrock encountered at the site was said to be typical of the upper layers of the bedrock encountered in excavations within Hawkesbury sandstone. The report then noted: The properties of the Hawkesbury Sandstone geological formation are well known to design engineers throughout the Sydney region, with many excavations and building works being undertaken in the rock strata. Further, whilst the rock is in its unweathered state is a relatively high strength rock and capable of standing at near vertical slopes in open excavations, in the more weathered states, the strength controlling features of the rock are its joints, seams and bedding planes (usually termed 'geological discontinuities'). Whilst the discussion of possible rock failure mechanism is beyond the scope of this report, it is considered obvious that if intersecting near vertical joints are present with rock bedding planes that dipped towards an excavation ... then such an excavation has a greater chance of failure than one in which the bedding planes do not dip towards the excavation. Thus, and by reference to the joint rosette on drawing No BA006G1 is it apparent that the eastern wall excavation of the cofferdam was more likely to be unstable than the north wall. The import of the foregoing is that at the time of preparing the design for the sheet pile cofferdam the designer would have been aware of: (a) the probability that a weathered sandstone bedrock into which the sheet piling was to be driven would have a number of geological discontinuities; (b) an important design consideration would be the orientation of geological discontinuities with respect to the planned excavation; (c) the usual geological situation that rivers are frequently associated with weaker underlying bedrock strata than the surrounding area ; as such, poorer quality bedrock strata could be expected as the Cooks River was approached;. (d) there was a distinct probability that an eastwards excavation towards a north/south face (ie the east wall) would be more unstable than the other excavation faces; (e) when an excavation face was created within the 'zone of influence' of the driven sheet piling, bedding planes and other geological discontinuities in the rock mass would be the controlling influence on the strength of the rock structure; (f) it would not be reasonable to rely on the 'intact rock mass strength' for the design of 'toe support' (or 'toe restraint') for a laterally loaded pile structure. 53 Mr Shirley understood that the defendant Coffey Partners was to provide the geotechnical parameters to be used in the design by Connell Wagner. A review of the parameters provided by Coffey Partners indicated "that they are reasonable and compatible with local experience of the Hawkesbury sandstone". It was noted that the parameters were preferably to be supplied in writing and that, whilst the parameters discussed and agreed in April 1997 appeared reasonable, it also appeared that Connell Wagner used parameters approximately twice those discussed in their design calculations. Further, whilst sheer strength of a soil/rock material was usually about one-half of the UCS such a correlation to a uniform intact material had no relevance to a weathered jointed and seamed rock strata. 54 Mr Shirley stated that the design concept of heavy duty steel boxed piling to be driven into moderately weathered sandstone was inherently sound when excavations were contemplated in the vicinity of the driven steel piling and there were many other design issues arising. Some of the matters that needed to be considered was the likely occurrence of clay/shale seams within the weathered bedrock below the piles, the effect of driving the sheet piling into the rock being radial cracking of the bedrock due to the stresses induced by pile driving, and rock joint openings due to lateral loadings from the sheet piles on the rock with consequent water in flow and sealing issues. This was particularly an issue at the southern end of the project. In view of these matters, Mr Shirley considered from the outset that the design concept for the sheet piling should have included anchors close to the toe of the box and sheet piles with such anchors being required to have been installed and stressed prior to any excavation within the zone of influence of the sheet pile. 55 Mr Shirley also regarded the width of the berm in front of the sheet pile wall as a critical issue. This was because the closer the excavation came to the sheet piling the more important it was that the sandstone in front of the wall could resist the required loads and provide the necessary toe restraint to the sheet piling. He calculated that a minimum horizontal berm width was about 3 to 4 metres although on an alternative approach a berm width of 2.4 to 3.4 metres might be used. A berm width of less than 2.4 to 3 metres could not be justified in the absence of effective supporting anchors at the base of the box pile. Therefore the design berm width of 500mm always required toe anchors for stability of the sheet pile wall. 56 Professor Carter's report on the strength of the sandstone was considered by Mr Shirley who noted that the rock surrounding the failed box pile was very weak (about 2.5 MPa), that whilst the box pile was founded on solid sandstone it had only been driven through weak/very weak sandstone and that there were cracks in the bedrock induced as a result of pile driving operations. Those observations, combined with the observations of Mr Shirley, confirmed that it was always essential to have a substantial bench in front of the sheet piling until suitable toe support anchors were installed and were effective. The pile driving caused new cracking in the bedrock with resultant weakening. 57 The cause of the failure, according to Mr Shirley, was the likelihood that one of the box piles was not driven to sufficient depth to allow for adequate pile toe restraint in the HAW2 sandstone and this lack of harder bedrock penetration could have initiated the failure of an already overtaxed support system. He reiterated his view that, without effective toe restraint anchors, the stability of the sheet piling could not be assured unless there was an effectively horizontal berm of at least 2.4 metres wide and a slope from the berm to the floor of the excavation. Further, the berm layout depicted on a Transfield Constructions Pty Ltd sketch of 30 October 1997 had an inadequate margin of safety against failure. Mr Shirley noted that, whilst the conclusions in his report were based on significant data, they were arrived at without the benefit of suitable full technical discussions with the defendant Coffey Partners or with Transfield Constructions Pty Ltd, Connell Wager and the workmen employed on the job at the time of the failure. In those circumstances, there could be a number of matters and circumstances of which the author was not aware and which could affect some of the conclusions. 58 In oral evidence, Mr Shirley said that weathering, rotting or deterioration occurring in the rock was the result of a soil layer. Gradually, the stone becomes less and less weathered with depth. Typically, Hawkesbury sandstone might be very weathered at the top but good rock would be found at about 6 metres depth - it became stronger unless weathered with depth. In the pile driving process, the surrounding rock, to a limited extent, might be affected and crack. It would be doubtful that rock more than 5 or 6 metres away would be affected. That is what was referred to as the "zone of influence". 59 In cross-examination, Mr Shirley said that he understood that the cofferdam was designed by Connell Wagner and that the defendant Coffey Partners had provided certain geotechnical parameters through Dr Redman. Mr Shirley knew of Dr Redman and that he was held in high regard for his expertise as a geotechnical engineer. 60 Mr Shirley was not aware that Connell Wagner had, in fact, modelled the design on the assumption that at least at a certain level of the excavation there would be no toe fixity. Mr Shirley believed that, in carrying out a design, a number of design scenarios would be considered including factoring in the weakness and fractures in the rock and how that might occur. He also understood that the pile excavation geometry of the berm at the eastern wall was 500mm but that had not yet occurred before the wall failed. He understood that there would be inspections of the berm before it was excavated to final geometry and, until that stage was reached, there would a substantial berm left. Mr Shirley was of the view that the sheet piling was underdesigned having regard to the parameters supplied - there was a component of the design on his calculations which did not conform with the recommended design parameters and was grossly overstressed. 61 From his own experience and from Professor Carter's report, it was obvious that driving a large steel box pile into rock would crack it. It was accepted by Mr Shirley that the only way box pile could be driven into a rock was by cracking it - the piling cracks the rock until it reaches a point where the rock is sufficiently hard and strong that it will not crack and that is where the pile terminates. Wherever sheet piling or a box pile is driven into rock there will be cracking of the rock and that was inevitable. Anyone designing a sheet pile structure in the case of a cofferdam would certainly be aware of that effect. 62 Mr Shirley clarified that if the berm had been comprised of soil materials, then it would have made an insignificant difference to the support it gave the walls. His view of a berm of 3 to 4 metres was based on a berm comprised of weathered rock strata and that would have made a substantial difference. 63 Mr Shirley was of the view that it would be virtually impossible to drive these piles very far into HAW2 sandstone - they could not have been driven 1.5 metres into HAW2 sandstone. He accepted that piles could be driven 300mm into HAW2. He did agree that if there was a 3.5 berm then it was approaching the situation where the wall was unlikely to fail. He was more comfortable with a 4 metre berm although a 3.5 berm with a sloping surface to the base of the other excavation would have avoided failure of the wall. The purpose of the sloping surface was to guard against "daylighting" of clay seams that may exist in the stronger rock or any unfavourable joints that might exist. EVIDENCE OF DR REDMAN 64 In November 1998, Dr Redman supplied a number of answers to questions posed by WorkCover and attached a number of documents relating to the work performed by the defendant at the Cahill Park site. In relation to that documentation, it was pointed out that, in May 1998, Connell Wagner had been informed by the defendant that reference should be made to individual bore logs and piezocone interpretation in assessing conditions at particular locations and that, in December 1996, in making recommendations to Connell Wagner, the defendant had stated that reference should be made to the long section and individual bore hole data, noting in particular "the range of rock strengths expected to be encountered when driving sheet piles". During the tender stage, the defendant had warned Connell Wagner that the design of sheet pile walling was geotechnically complex and offered to assist in the design. That offer was not taken up. Dr Redman had not at any time seen design calculations performed by Connell Wagner for the project. The defendant had never been engaged for the detailed design phase of the cofferdam project. 65 According to Dr Redman, Connell Wagner were the designers of the cofferdam and the defendant provided input with regard to the geotechnical model and parameters. The role of the defendant in design did not extend beyond comment at the concept design stage. It was accepted that in providing advice in regard to the geotechnical model and parameters, the defendant had occasion to carry out onsite visits and provide other advice on an "as requested" basis. At no time was the defendant provided with final drawings showing the support designed for the eastern wall. Further the defendant was not asked to provide corresponding comment on any Wallop computer models or other design calculations for the cofferdam design at Cahill Park. 66 In relation to the geotechnical investigations carried out at Cahill Park, the following were included: (a) cored and non-cored boreholes; (b) piezocone probes; (c) field mapping of rock exposures in nearby cuttings; (d) installation of stand pipe piezo meters; (e) down hole water pressure testing; (f) slug permeability testing of piezometers; (g) seismic refraction survey; (h) tomographic imaging of interpreted rock profiles for site; (i) geological mapping of cliff faces on the Princes Highway at Arncliffe; (j) laboratory testing of soil and rock samples. Dr Redman said that the scope of work was assessed on the basis of providing a sub-surface model that could be used in conjunction with supplementary drilling, test pitting and site observation during the construction stage. 67 The use of the categories HAW3 and HAW2 was described by Dr Redman by reference to well known rock classes referred to in Pels et al in a 1978 document entitled "Design loadings or foundations on shale and sandstone in the Sydney region". In summary, HAW3 was Hawkesbury sandstone class V under the Pels classification, HAW2 was Hawkesbury sandstone class III and IV under that system. 68 The investigation carried out for the construction of the underground rail link by the defendant involved a number of regional studies of bedrock elevation that included the region between Redfern and Bulli Creek. Dr Redman said the investigations included regional magnetics down hole and surface seismic refraction studies. These investigations, together with more site specific geophysics and field mapping, indicated an expected variation in elevation and character of the bedrock profile. It was pointed out that there were 12 boreholes in the 300 metre section at and about the location of the failure. The contours and rock levels at nominated places showed that the rock/alluvium contact sloped down gently to the east of the location of the eastern wall and that these levels were confirmed during construction from piling records and test pits A and B. This had also been further confirmed by drilling completed since the failure of the eastern wall. 69 Dr Redman rejected the suggestion that HAW2 and HAW3 sandstone were "unreliable". Having regard to the Pels classification, the distinction between class V (HAW3) and classes III and IV (HAW2) would, it was asserted in most circumstances, be quite clear: the definition identified a recognisable level of a range of strengths and defects. The expected variability in strata level was the reason for carrying out extensive seismic refraction and tomographic imaging. The seismic imaging represented a state of the art approach to use and interpretation of geophysics data and was well in advance of general geotechnical practice in Australia as well as elsewhere in the world. 70 Implicit in the development of the overall scope of the investigation and assessment was the fact that the piling records would provide information on variability and that, as part of the design concept, there would be a requirement for inspection of rock quality and an assessment of whether there was need for support to be provided. Dr Redman said that the nature of some surface contacts was such that, generally, they would not be fully defined unless and until they were exposed during construction. Prior to exposure there would always be uncertainty regardless of the number of boreholes drilled: design inspection and construction procedures had to allow for this prospect. Indeed, investigations carried out since October 1997 involving many further boreholes indicated conditions were consistent with the long section model developed prior to the project. The distinction between HAW2 and HAW3 appeared to be irrelevant because testing after the failure of the eastern wall indicated that the rock mass at the toe of the wall would have had a strength greater than HAW3 in any event. 71 There were four steps involved in achieving correct embedment in the rock the first of which was the development of an accurate or conservative geotechnical model. Dr Redman said that the defendant had completed this step and was involved with Connell Wagner in confirming geotechnical conditions by inspections at appropriate times to ensure that the conditions were at least as good as assumed in the design. Dr Redman stated that only the designer fully knew the assumptions relied on in its design and those assumptions needed to be verified by inspection and other means during construction. The defendant understood that the design at final excavation had not originally allowed for the existence of a defect in the rock at the toe and so the inspection regime had to verify that such defects did not exist prior to such defect threatening the safety of the structure. 72 The defendant did not determine "toe fixity": it carried out site visits to observe the rock at various levels to clarify it as HAW2 or HAW3 and to assess that, it, amongst other things, had an undrained sheer strength of 500 kPa Site visits also checked for the existence of adversely orientated and localised defects. HAW3, corresponding to Class V sandstone, was of a strength that could be crumbled by hand whereas HAW2 was a Class III/IV sandstone. Dr Redman said that the 500 kPa undrained sheer strength value was assumed by the defendant as being one that the designer would use in conjunction with the planned berm width to arrive at an ultimate lateral resistance value. It was presumed that if more than one value of toe fixity had been used in the design then the designer on his own assessment of the borehole data would chose to use a different value of strength. Notwithstanding this possibility, the designer would still need to assess the support requirements for the temporary or inspection stage based on the assumption that the expectations regarding rock strength may not be able to be relied upon due to the presence of local structure or any variability, stating that some form of additional toe support before final excavation could take place. 73 On 3 October 1997, the defendant, following general observations of the cofferdam structure, raised a number of concerns about the overall structure with Connell Wagner. In a facsimile confirming the conversations, Connell Wagner were advised, amongst other things, to avoid excavation of the rock near the sheet piles until Connell Wagner was confident that toe support was sufficient for the loads currently being experienced. 74 Two representatives of the defendant (Dr Redman and Mr Parker) inspected the site on 21 October 1997 for the purpose of observing the conditions exposed in excavated faces near the services area and to arrive at recommendations for any shotcreting lagging or other support to the exposed faces. During that visit they became aware that shotcrete had peeled from some sheet piling on the southern face near the services area. The defendant had specified the length and spacing of dowels to be installed as part of the shotcreting, but inspection that day revealed that the dowels had not been installed at more than one location. It was established that there were no other areas where circumstances existed similar to that of the southern face near the services area where the shotcrete had peeled. While on this site visit, Transfield Constructions Pty Ltd and Connell Wagner were advised of the defendant's assessment and recommendations. In relation to the east face, Dr Redman, by facsimile sent on the same day, had advised that toe support was to be provided before any further excavation or removal of material. Connell Wagner had been asked to comment on the possible frequency and installation sequence for the toe support, although it was advised that the sequence should involve opening up the area of one pile only at a time or in accordance with Connell Wagner's sequence, provided it was assumed that toe support could not be generally relied upon until an anchor bolt was in place. Once toe support was in place, shotcrete was to be applied. The recommendations presumed that there was adequate treatment of any localised water pressure or leakage from the clutch piles or for other reasons and it was assumed that grouting and/or welding of plates or that concrete plugs would be used. 75 Dr Redman said that he had been asked to assist in a presentation to workers and to explain the assessment of the shotcrete problems that had occurred at the southern wall. Mr Di Cesare of Transfield Constructions Pty Ltd had addressed workers in the lunch room and Dr Redman said he explained the assessment of the shotcrete mechanism, described the process of assessment that had been undertaken with the surveyor and said that he was satisfied that the shotcrete issue had been addressed: he also noted that there was a continuing programme of installation of toe anchors to the east. 76 The decision to recommend installation of toe anchors to the eastern wall was made by Dr Redman on the basis of the extent of rock structure being observed in the services area. He said that the engineering judgement was made that circumstances were such that it was appropriate to install anchors rather than awaiting an inspection stage to assess rock quality. The decision to install anchors was an alternative to the inspection stage and simplified the process. Dr Redman said that at all times it was presumed that the structure would have been designed to be safe: if the design was suitable at the time at which an inspection was to be made and construction had not proceeded past the appropriate inspection point, the structure should have been safe at that time. 77 When asked why the defendant had not advised Transfield Constructions Pty Ltd to suspend work inside the cofferdam, Dr Redman answered that the defendant had fully assessed the shotcrete issue and was satisfied that risk did not exist to the workmen from the further peeling off of shotcrete. The defendant had observed exposures of rock in the services area and recommendations were made for further lagging and shotcreting works to be carried out. Rock exposures at the eastern end of the Cahill Park section were largely obscured by the berm debris and sand inflow. It was presumed that the area had not been excavated beyond inspection stage and would be consistent with the circumstances assessed by Connell Wagner in response to the defendant's facsimile of 3 October 1997. At no time did personnel from the defendant have information that could lead to a belief that there was an unacceptable risk of failure at the eastern end. The defendant reasonably expected that its advice to the designer had been appropriately followed. The previous concern raised about Cahill Park with Connell Wagner was confirmed by the 3 October 1997 facsimile and discussion concerning support being provided at the project. It was to be noted that the facsimile referred to excavation of rock near sheet piles until Connell Wagner were confident that the toe support was sufficient for the loads being experienced. Dr Redman felt that this was a very explicit and direct expression of concern, and that Connell Wagner had not advised the defendant there were any problems or concerns. Connell Wagner had been present on site on the weekend prior to the failure and the defendant did not see its role to be second-guessing the designer in the absence of any design review and having already expressed its concerns in the facsimile of 3 October 19987. Dr Redman noted that since the failure of the east wall the defendant's inspection indicated that the rock mass at the toe of the eastern wall, would have had an undrained sheer strength in excess of 500 kPa. Thus either the design required more strength or construction had removed more rock. The defendant did not believe that there was a defect at the base of the sheet pile although it was possible that one did exist, then the design had not properly considered that occurrence or excavation had proceeded too far. 78 In oral evidence, Mr Redman said that, while he was on site on 21 October 1997 just before the collapse of the eastern wall, he was unable to tell if the berm was excavated 1.5 metres above the toe of the sheet piling because the piling was not marked as he had advised on 17 September 1997. 79 From Friday 17 October 1997, the defendant through Dr Redman and Mr Parker was on site every day. Rather than weekly visits, this daily presence had occurred because of the shotcrete issue, the water inflow at the northern wall and remedial measures being considered, and also to permit an inspection. On 20 and 21 October 1997, visits were undertaken to observe the rock on the northern and southern walls and to assess the likely conditions at Cahill Park. On Tuesday 21 October 1997, apart from assessing Cahill Park, Dr Redman also addressed the workers. 80 Dr Redman noted there were many documents passing between Transfield Constructions Pty Ltd and Connell Wagner that had not been sent to the defendant. The defendant had a limited role but, nevertheless, had strongly recommended against over excavation and had advised on a variety of issues as they arose. By reference to site notes, facsimiles and other documents, Dr Redman had compiled a chronology of events relating to the project. In late May 1997 there had been a recommendation to Connell Wagner to design a restraint to the southern wall toe as soon as possible, and a test pit had been dug to locate the rock. In mid-July 1997, test pits were excavated to assess rock levels under the highway and low strength sandstone levels were picked up by the surveyor. The defendant advised that further excavation could take place there but limiting the excavation to low strength rock level when another assessment could be made by excavation of pits or a trench at least 2 metres away from the line of the sheet piling. On 18 August 1997, it was noted that excavation of the northern face under the highway had proceeded to a level of shale with only 0.5 metre berm left - this was contrary to previous recommendations and the defendant advised backfilling against the face of the wall. On 17 September 1997, there was a slump below the surface area. It was noted that Transfield Constructions Pty Ltd had admitted that it had erred by over-excavating the face contrary to the defendant's advice. In relation to the eastern end of Cahill Park, Transfield Constructions Pty Ltd had proceeded with excavation contrary to the defendant's advice and the defendant instructed it to immediately stop the excavation and to reinstate that area. It was recommended that excavation be limited to 1.5 metres above the toe of the sheet piling, identifying the level on the sheet piles together with those identified by Connell Wagner as requiring toe anchors. On 22 September 1997, water was observed coming from beneath sheet pile No P30 and the grouting contractor was moved over to deal with that leak. On 29 September 1997, Mr Di Cesare telephoned the defendant saying that they were experiencing difficulty installing anchors because of the water flow. Transfield Constructions Pty Ltd said that it could live with a 2 metre wide berm in the long term on the eastern face. The next day, two test pits were dug at Cahill Park to obtain data on the rock depth. On 1 October 1997, the defendant advised Connell Wagner that a 3 metre berm should be left and that an assessment should then take place of rock quality and whether rock anchors were required. On 3 October 1997, the defendant sent a facsimile to Connell Wagner confirming earlier discussions and advising, amongst other things, to avoid excavation of rock near the sheet piles until Connell Wagner was confident that the toe support was sufficient for the loads currently being experienced. On 18 October 1997, it appeared that the block of rock in front of the sheet piles between pile P8 and P10 had moved into the excavation, and water was coming in from an extremely weathered seam about 1 metre below the berm level. The notes recalled an agreement that temporary strutting should be installed. On 19 October 1997, water flow was noted as occurring near pile P10 and at piles P21 and P22. There was, however, "a rapid drop off" with tidal influences and a general decrease in the overall flow rate. On 21 October 1997, the problem with falling shotcrete was noted. A number of discussions were held and a general comment was made to Mr Di Cesare that work had to proceed cautiously, the sequence of bolting had to be sorted out with Connell Wagner first and they should proceed panel by panel without broad scale excavation or removal of material. 81 Dr Redman said that on 3 October 1997 he had become formally aware of problems with strut loads and reported de-stressing of anchors, which were supposed to be tension anchors on the northern face. Although nothing was actually said formally to him, he was a little troubled that these things were going on and, whilst he did not know the details of the monitoring referred to, he took the opportunity of looking around Cahill Park to try to understand whether it was possible to find out what these things might be. There were various reasons a stressed anchor might de-stress, and there were reasons why a strut might lose load, but the particular explanation could be of concern. He concluded after looking around the site that he could not understand what was actually occurring and resolved to tell Connell Wagner that he did not like what he was hearing and that they should investigate it and sort it out before any other work took place. 82 On 18 October 1997, Dr Redman had spent considerable time at the site because of Transfield Constructions Pty Ltd's concerns. He recalled a meeting with Mr O'Brien of Connell Wagner and Mr Shepherd in circumstances where, on the southern wall, shotcrete had peeled off and as a consequence several pile toes had been exposed. Later in the evening, there was water ingress on the northern side almost directly opposite. In the context of problems on the northern and southern walls, Mr Shepherd asked if there was a likelihood of a catastrophic failure of the dam and, having been satisfied that the pile on the northern wall had not moved although there had been a water in flow, he concluded that a catastrophic failure of the northern and southern walls was unlikely. During this conversation, no issue was raised by Mr Shepherd about the eastern wall. Dr Redman had asked if there were any problems and had been informed that there was a de-clutching problem at the eastern wall but that it was being attended to by the grouting company. He walked past that area while inspecting the site. 83 At this time, Dr Redman did not ask whether there had been excavation past 1.5 metres above the toe of the piles leaving a berm of at least 3 metres. He had already addressed those concerns in early October with Connell Wagner and he was comfortable that those matters were being addressed. On 18 October 1997, he was focusing on the southern and northern walls. 84 On 19 October 1997, there were no new problems and he attended the site to observe the installation of the temporary strutting that had been proposed. On 21 October 1997, there had still been no inspection at the eastern wall but because of his observations in relation to the northern and southern walls, Dr Redman recommended that there should be rock bolts installed and that there should be no further excavation until that took place. It was no part of the defendant's role to nominate which box pile should be bolted first or their sequence: it was up to Connell Wagner as the designer of the wall, who knew what the loads were, to work out how it wished to go about that process. 85 When Dr Redman addressed the men in the shed on 21 October 1997, and when he had made recommendations to Connell Wagner, he had assumed that the design of the cofferdam had an adequate factor of safety when there was a berm in place of 3.5 metres wide and 1.5 metres above the toe. The defendant had no reason to doubt that the designer had appropriately considered the design. On 21 October 1997, before Dr Redman spoke to the men in the shed, he had taken no measurements at the eastern wall. At that time, the defendant had no information that could lead to a belief that there was an unacceptable risk of failure at the eastern end. Dr Redman added that the visits that took place on 20 and 21 October 1997 took advantage of the exposures of rock which were then available on the northern and southern face in order to address what should be done at the eastern wall. Dr Redman rejected the proposition that he and Mr Parker (also from the defendant) were on site from Saturday 18 October 1997, because it was clear that they were required to advise as to the safety of the site within their area of expertise. 86 The berm in front of the eastern wall was obscured by spoil and other material during the time that Dr Redman was on site but he thought that was of no real relevance, in that his role at the time was to assess whether or not there was a need to advise the use of rock anchors or some other support to the eastern wall. He had not asked anybody else, including Transfield Constructions Pty Ltd, about the size of the berm on the day that he spoke to the men. There was a presumption that both Transfield Constructions Pty Ltd and the designer of the dam Connell Wagner had done their job properly. That presumption itself was not based upon anything that was said to him by representatives of Transfield Constructions Pty Ltd or Connell Wagner. On 17 September 1997, although it was not directly part of the defendant's brief, he had raised objections in strong terms to over-excavation near the eastern wall. It was potentially a safety issue and contrary to advice that had been given to Connell Wagner. Dr Redman said that he had told Mr Shepherd that there needed to be a focus on the requirement for berms. Mr Shepherd undertook to address that issue and Dr Redman no longer had a concern being satisfied that the incident had been addressed. He felt no necessity, having had the agreement of the project manager, for raising the issue again between 17 September 1997 and 21 October 1997. 87 On the information available to him, Dr Redman said that he was satisfied that there was no risk of the eastern wall failing on 21 October 1997. Had he known of what he now believed to be flaws in the design of the cofferdam he would not have held the same view. There was no need to have the men withdrawn from the site while the rock bolting was happening because that was a step that had been taken all over the site as part of the usual process of securing the wall of the cofferdam. By recommending the installation of rock bolts, Dr Redman stated that he was not identifying a risk to the security of the wall and therefore the safety of those employed to work in the dam. The decision to rock bolt was one taken in the normal course of the project assuming that the design was adequate and to ensure that after final excavation that support remained for the wall. 88 In cross-examination, Dr Redman said that the borehole sunk closest to the eastern wall was some 10 metres from it and it showed an unconfined compressive strength of 2 MPa. 89 In a March 1996 meeting at the offices of Transfield Constructions Pty Ltd where a representative of Connell Wagner was present, Dr Redman had explained that driving sheet piles into HAW2 sandstone could be difficult and the issue, geotechnically, considering the nature of the Cooks River Crossing, was of water ingress with the sheet piles acting as a type of water cut off through the river. The sheet piles were targeted for penetration into the rock so as to achieve a foundation rather than having them rest in soft estuarine clays. This served a purpose also of cutting off water flowing underneath the shot pile into the excavation. It was theoretically possible to rest the sheet piling on the rock rather than embedding it in the rock but, in that circumstance, the rock would be serving no function in terms of retaining the toe in any particular position other than in the vertical direction and would have nothing to do with lateral resistance. 90 Connell Wagner had previously been supplied with information regarding another aspect of the work and that geotechnical information was confirmed. It was suggested that Connell Wagner should look at individual bore hole logs and piezocones for conditions of particular locations. Dr Redman expected that Connell Wagner would consider the information provided. Up until May 1997, Dr Redman was not aware that the designer required advice in relation to quantifying restraint and, when that advice was sought, it appeared that some allowance was to be made for the use of rock for lateral toe restraint. 91 Connell Wagner had a number of design options available for this temporary work at Cahill Park. After the concept stage, detailed design commenced when a particular way of doing the work was selected: in this case, the use of sheet piles required consideration of the props required and where they were required and how big they needed to be and the loads for which they were to be designed. Dr Redman stated that it was not necessary for the geotechnical adviser to know the design detail: the geotechnical engineer provided a model of the ground, which the designer used for its proposal. In this case there had been a meeting in December 1996 at the offices of Transfield Constructions Pty Ltd where the temporary works were discussed. The discussion was held in the context of sheet piles being founded on the top of HAW2 for vertical load and in assessing the scale of any bench to be left for maintaining the integrity of the excavated face. At this stage, there was no suggestion that sheet pile would actually penetrate the rock although Dr Redman had earlier supplied advice about how far sheet piles could be driven into the rock. In the concept of driving sheet piles to form a bench, that approach meant that it was necessary to inspect the bench to ensure that the rock was adequate, that there were no adverse defects and whether the use of shotcrete might be necessary to ensure that parts of the face did not come away or fall out. At this meeting, Dr Redman prepared a sketch of what was being considered which showed sheet piling meeting a bench: the sketch contained a note that inspection was needed to assess the anchoring requirements. 92 Transfield Constructions Pty Ltd and Connell Wagner were involved in a relatively complicated structural feature in that the structural designer had a number of options about the way in which to address problems as they arose. It was in that context that Dr Redman supplied advice when requested rather than asking questions about what they were proposing to do: he understood that the designer had a number of options available. 93 In May 1997, Dr Redman first gave advice in relation to a parameter for use in assessing lateral restraint or lateral resistance and, being conservative, gave a parameter that actually downgraded a little the strength and value of the sandstone. HAW2 was by definition Class III rock with Class IV being the lowest having its lowest strength range to 2 MPa. At this stage, nor indeed at any other stage, was Dr Redman informed of the actual forces being resisted at the toe. He had not known the processes that Connell Wagner had gone through, he had not seen any of their design calculations nor a description of their methodology. Dr Redman did not believe, acting professionally as a consultant in this matter, he was required to ask questions about the design - much depended on how much the designer chose to do himself and what advice the designer then sought. It was not unusual in Dr Redman's experience for large consultants like Connell Wagner to be provided with geotechnical parameters and then not to hear from them again. The level of interaction was a matter for the designer. 94 A Connell Wagner document relating to Cahill Park and dated 30 June 1997 was regarded by Dr Redman as presenting a no toe fixity approach to the project. The designer appeared to assume that any contribution being made to the toe from the HAW2 rock was less than had been recommended and spoke of that in the context of the presence of blocks or wedges or jointed sandstone. Dr Redman was not sure whether the document meant that there would be no support whatsoever because a temporary berm would be providing some restraint. This no toe fixity scenario was essentially a very poor rock quality scenario. The designer appeared to assume that there would not be any significant lateral restraint at the toe. It appeared that the design of the cofferdam would be on the basis that the restraint obtained at the toe would be from the soil berm and not from the HAW2 at 50 kPa. 95 It appeared to Dr Redman that Connell Wagner were designing for the use of a berm at least and were making provision for an inspection stage. The purpose of the inspection stage was to ascertain whether there was lateral support or whether it was sufficient. A berm may be assumed to provide some support and, if the quality of rock is poor, the wall is still safe because of the assumption of the berm. If the rock quality is sufficient then the berm can be removed or trimmed back. If the rock quality is poor anchors could be installed. Once the anchors are in place the berm can be trimmed. The June 1997 Connell Wagner document seemed to assume that there was poor quality rock at the toe and that there was to be a step reached when monitoring of pile movements and/or loads would indicate that there was poor quality rock at the toe and that therefore it would be possible to take action and install a waler to provide further support in that situation. At the time, Dr Redman had no information about how the load was being measured although he had since learned that load cell monitoring equipment had been used. Dr Redman had not seen any load cell readings prior to October 1997. He had simply heard an informal comment that there were problems with load cell readings. That is when he spoke to Connell Wagner on 3 October 1997. If he had been aware of the report from Hercules Engineering, that the load on the walls was very significant and to take precautions and checks accordingly, he would have been alerted to raise the issue with Connell Wagner because something was happening at the site. If there was a problem with strange readings from the load cells it would have been appropriate for Transfield Constructions Pty Ltd to stop work and for remedial measures to be addressed. Prior to 22 October 1997, Dr Redman was unaware of the precise terms of the Joint Ventures concerns relating to toe fixity of the sheet piling at Cahill Park. In April 1997, he had spoken to Mr Shepherd who indicated that some issue had been raised and Dr Redman spoke of the role of the inspection regime, as the issue seemed to relate to addressing the quality of the rock. Dr Redman understood that, as a result of that conversation, the issue was resolved. As he understood it, the Joint Venture had raised issues about an aspect of reliance on the toe pitch and he had reinforced the advice that had been previously given, that the berm would have to be inspected prior to being trimmed to final geometry to assess the nature of defects if present and to take appropriate action at that stage. 96 In relation to the borehole closest to the eastern wall, Dr Redman would have derived a value of 2 MP or perhaps an even lower value, because the Connell Wagner document of 30 June 1997 considered circumstances where the rock was of poor quality or jointed. There were areas of poor quality and some near the borehole closest to the eastern wall which were poorer quality than 2 MPa. 97 On 19 August 1997, Dr Redman had advised Connell Wagner by facsimile confirming advice given the day before that excavation on the northern side of the works was contrary to previous recommendations, "to proceed cautiously by leaving a three to four metre bench and to progressively anchor the toe of the sheet piles before excavating the final bench configuration". Dr Redman said that he had spoken to Mr Di Cesare on site about his concern and Mr Di Cesare had asked if they had "done wrong?", to which Dr Redman had replied "Absolutely". Dr Redman explained what had been the agreed procedure of leaving the bench and that he had previously given that advice. He also instructed that there be backfilling at the point of excavation. That direction was complied with and two anchors were subsequently installed. 98 On 17 September 1997, Dr Redman had sent a facsimile to Transfield Constructions Pty Ltd which, amongst other things, dealt with the eastern end of Cahill Park. In that facsimile, he advised that excavation was to be limited to 1.5 metres above the toe of the sheet piling and that this level was to be identified on the sheet pile together with those sheet piles to be anchored. Installation of the toe anchors was to take place progressively as previously advised. Dr Redman also stated in this facsimile that excavation to a low level should not proceed until the anchors were installed unless a 4 metre wide berm was left in place adjacent to the sheet pile. 99 On 4 September 1997, Transfield Constructions Pty Ltd entities held a meeting where it was reported that the Joint Venture had concerns over toe restraint in the cofferdam. It also recorded that load cells had been installed to struts in Cahill Park and that readings to date indicated that loadings were within expectations. Dr Redman did not attend that meeting nor was he advised of those concerns. As these directions by the defendant were given in August and September 1997, Dr Redman had no reason to believe that his instructions had not been followed. 100 An internal Connell Wagner memorandum dated 19 October 1997 had not, prior to 22 October 1997, been seen by Dr Redman. That document noted that, in relation to sheet piling on the Cooks River side immediately west of the Princess Highway Bridge, the sheet piles were driven into very soft rock and apparently terminated at the level of a horizontal joint plane. The document went on to state that toe failure could occur and rock anchors were to be installed in each sheet pile as soon as possible to prevent a failure occurring. This memorandum recorded the same type of topics that Dr Redman had advised in relation to the project. 101 On 30 September 1997, Dr Redman had attended a meeting with Mr Shepherd and Mr Di Cesare and discussed problems that Transfield Constructions Pty Ltd were having installing rock bolts because of the width of the berm. There was a suggestion from Transfield Constructions Pty Ltd that the berm be reduced. Dr Redman looked at driving records and test pits and decided that rock levels were probably more consistent than might be interpreted from a strict application of the driving criteria. There was a possibility that the assessment, based on the driving criteria as to which piles needed rock bolting, could be re-visited and that it was preferable to carry out an inspection when there was a berm present in front of the toe to assess the conditions. The berm was to be 3 metres. In giving that advice he had formed the view that the rock would be likely to be HAW2 because there had not been any confirmation yet that this was HAW2 level rock. Dr Redman interpreted several pieces of information. The proposal was that there could be a re-assessment both in terms of level, quality of rock and whether or not there were any adverse defects present. It was possible that after re-assessment rock anchors might be used. 102 On 3 October 1997, following a discussion on site, Dr Redman had sent a facsimile to Connell Wagner. In that facsimile, he advised that excavation of rock near sheet piles was to be avoided until Connell Wagner was convinced that the toe support was adequate for the loads currently being experienced. Dr Redman said that later he received a telephone call and was advised that the problems with the load cells had been sorted out: he had raised some concerns and they had dealt with the issue. 103 When on site on 18 October 1997, Dr Redman said that he noticed a company called Ground Consolidation doing some grouting work near the centre of the eastern wall. He said that at times gaps could occur in sheet piles after they were driven and water could come through which was not seen until excavation reached a particular level. This was dealt with by grouting to seal the gap. As he was passing this work he asked Mr Shepherd what was going on and Mr Shepherd said that it was merely grouting of the area where there had been some declutching. Mr Shepherd had not initiated that conversation. Dr Redman denied that he said to Mr Shepherd that it was more a construction nuisance issue to be dealt with by pumping. During that visit water had come in from both the southern side and the northern side and Dr Redman advised Mr Shepherd that it would be prudent to install some temporary support. Mr Shepherd had asked him whether the dam was likely to catastrophically fail as the sheet pile on the northern site had not moved. Dr Redman denied telling Mr Shepherd that the dam was "structurally sound". 104 On 22 October 1997, Dr Redman was unaware that the excavator driver, Mr Galloway, had observed a 2 metre berm at the eastern wall. He was not aware that the berm had been reduced to this extent and, if he had known, he would have had the men removed from the site while further inspection was undertaken. There was meant to be a wider berm until the rock bolts were installed. He had given no advice that the berm at the eastern wall could be reduced to 2 metres. 105 On 24 October 1997, a meeting was held at the offices of Transfield Constructions Pty Ltd which Dr Redman attended together with a number of other people including Mr Di Cesare and Mr Shepherd. Dr Redman recalled that during the meeting Mr Di Cesare said that the size of the berm at the eastern wall at the time of the failure was 2 to 2.5 metres. Dr Redman had informed the meeting that the berm was to be 3.5 metres. 106 When Dr Redman was on site on 21 October 1997, the berm at the eastern end was partly obscured with sand, disturbed material and excavation material on top of it. Dr Redman said he was unable to see how wide the berm was because it was to an extent obscured by debris. It may have been possible to excavate the material away in order to check the berm but he did not know how long that would take. It was not a concern to Dr Redman that, because of the spoil, he could not see the berm. The whole site, even to the north and south, had excavated material - he did not believe that the berm would be different at the eastern end. OTHER EVIDENCE 107 Mr Galloway was a plant operator and at the Cahill Park Project he was driving an excavator. In a statement given to Inspector Dubois Mr Galloway said that at the eastern end of the cofferdam there was a 2 metre wide bench in front of the cofferdam wall with about a 1.5 metre step down in front of it. 108 On the Saturday before the incident there had been a bad leak on the northern wall which stopped itself and had been pumped out. Mr Galloway went back to the site on Sunday morning and found there was "surcharging sand" coming in around the centre of the eastern wall and that leak continued. Just before the eastern wall of the dam collapsed on the Wednesday, Mr Galloway had seen Mr Mion, the Transfield Constructions Pty Ltd's engineer, Mr Di Cesare and one of the men from Ground Consolidation inspecting the wall. They passed no comment to Mr Galloway about the wall. 109 In oral evidence, Mr Galloway said that he had not measured the bench in front of the eastern wall but that 2 metres was his estimate of its width. Mr Galloway explained that, when he referred to water "surcharging" through the eastern wall, it was not clear water but was water bringing sand through with it all the time. It was a consistent leak. Loose material was used to create a dam in front of that leak to contain the sand that was coming in. Mr Galloway had been working for about a month at the site. When he got there the 2 metre bench had already been created but he had not excavated it further. He had never been told by anyone, including Mr Mion, of the dimension of the bench to be left in front of the eastern wall. On the Friday before the eastern wall collapsed he had not trimmed the bench in front of that wall although he had trimmed the northern and southern walls. Mr Galloway denied clearing material off the bench with his excavator and stated that, on 22 October 1997 (the day of the collapse) he could not recall placing an excavator bucket against the eastern wall to stop the flow of water. 110 Mr Galloway had raised some concerns about the shotcrete coming off the southern wall in particular, and he was aware of a safety meeting being called as a result. He did not attend any meeting of the men in the sheds in the few days before the eastern wall collapsed. 111 Mr Patton was employed at the site by Transfield Constructions Pty Ltd as a foreman. He was present at the site when the eastern wall gave way and the site flooded. He organised Emergency Services to attend, conducted a head count, saw two employees rescued from the water and attended to until the ambulance arrived. 112 He described a "massive influx of water mud and sand" coming in on the northern wall of Cahill Park on the previous Saturday. Connell Wagner and the respondent Coffey Partners were called to the site and later boilermakers were called to install horizontal braces across the cofferdam. They worked all Saturday night and early Sunday morning to install two braces. On the Monday morning, a section of shotcrete on the southern side of Cahill Park collapsed. After morning tea, the chairman of the safety committee approached Mr Patton saying that the workers were unhappy about the collapse and the use of additional braces and that they wanted an explanation from management. Mr Patton organised a meeting of the workers in the crib hut where they were addressed by Mr Di Cesare and Dr Redman from Coffey Partners. Dr Redman explained why the shotcrete collapsed on the southern side was not a concern and that the additional bracing was due to a broken clutch at the base of the piles on the northern side of the cofferdam. That is why there had been an influx of water and sand but, he said, it was perfectly alright to work in the dam. 113 On the next day, some cracks started appearing in the same shotcrete on the northern wall and he was asked by the acting project manager to install some steel bracing to support the shotcrete in case of collapse. On the following day when the eastern wall collapsed, Mr Patton said a 20 tonne excavator was trimming up the excavation to the eastern wall. At that time, there was another influx of water at the eastern wall which appeared to be coming from the base of the sheet pile. Mr Patton said the excavator operator placed his bucket over the incoming water to stem the flow. Prior to that, there had been a collapse of sandstone and dirt on the southern wall of the cofferdam at Cahill Park. 114 In oral evidence, Mr Patton said that events that he identified in his statement to the Inspector as happening on the Monday could have occurred on the Tuesday before the eastern wall collapsed. 115 At the meeting in the hut, Mr Di Cesare spoke only for a few minutes but he did the "main talking". Mr Patton regarded Dr Redman as addressing the whole of the cofferdam rather than issues relating to the northern and southern walls. He recalled Dr Redman explaining how the shotcrete had fallen, how there had been an excavation and how that occurrence was unlikely to be repeated. He had said that the site was safe and also said that anchors were to be installed at the toe of the sheet piles to the east. In cross-examination, Mr Patton could not recall seeing the excavator trimming the eastern wall on the day of the collapse and he may have been wrong when he put that in his statement when speaking to the WorkCover Inspector. Mr Patton was not sure whether it was the Saturday or the Tuesday when he saw the bucket being placed against the water coming in at the eastern wall. He had previously given evidence that this occurred on the day of the collapse of the eastern wall. Mr Patton accepted that he did not need to inform Mr Di Cesare that the excavator driver had placed a bucket against the leak at the eastern wall because Mr Di Cesare and Mr Mion were at the eastern wall. 116 Mr Mion was a foreman who had been employed to work on the Cahill Park site. He was involved in the rescue of the two workers swept away when the wall collapsed. Prior to that incident, he had no concerns about the safety of the site. 117 There had been problems on the site from the Saturday night when a major leak occurred and Mr Mion had constructed a "bung" out of sandstone to try to contain it. He had ordered larger pumps to cope with the amount of water coming in at the northern wall and noted that the specialist firm Ground Consolidated had been brought in to fix or control the leak. Those steps were partially successful in stopping the water although a leak occurred just off the centre of the eastern wall. At about that time, a rock shelf on the northern wall moved and he instructed the workmen to evacuate the site. During the course of Saturday night, Mr Mion saw Dr Redman at the site with another person from Coffey Partners. He participated in the work where struts were installed from the southern wall to the northern wall. He noted that during Sunday the water flows were steady with sand migrating into the area with the water flow. On Monday, there had been a large amount of concrete pumping between the sheet piling and rocks. On Tuesday, on the southern wall a large amount of sheet piling gave way which raised concern about the safety of the site. He recalled that at lunchtime Mr Di Cesare called a meeting of employees in the lunch room and told everyone that the design and the calculations had been reviewed and that the cofferdam was safe. After that, the employees went back to work. 118 In oral evidence, Mr Mion said he could only vaguely remember what was said in the meeting of the men and what was said by Dr Redman. He remembered statements being made about sheet piling being driven into Class 11 sandstone and that was considered to be safe. He regarded Class 11 sandstone as quite hard. It was his recollection that Mr Redman had said this, but there were other people who spoke about the state of the job as well. Although he could not remember precisely what Dr Redman said, it was something that made him consider it was safe to return to the site. Mr Mion said that he had left the bench of rock along the eastern wall that was approximately 3 metres wide by the width of the excavation - that is, across the full width of the excavation. The major leak had been at the northern wall but it was not a great volume of water coming in at the eastern wall. 119 In cross-examination, Mr Mion said that he took his directions at the site from Mr Di Cesare and Mr Shepherd. They had told him that there was a bench or berm of 3 - 4 metres to be left at a height of 1.5 metres in front of the sheet piling. He recalled an onsite meeting in September 1997 where Dr Redman was present and a direction that the piles be marked to ensure that 1.5 metres was kept, but he did not think there had been any over-excavation of the bench at that time. Mr Mion accepted that he had written a site note in September 1997 that excavation of the eastern wall had been stopped due to over-excavation but that was not in relation to the bench - too much material had been removed below the position where the anchors were supposed to have been installed. He did not recall Dr Redman swearing and saying that the over-excavation had to stop. 120 Mr Mion recalled being informed by either Mr Di Cesare or Mr Shepherd, in early October 1997, that at the eastern wall, a bench of 3.5 metres wide was to be left in front of each pile. He believed that instruction had been followed. 121 By reference to photographs, Mr Mion could not accept that the eastern end of the site was being excavated to the level referred to as "boxed out", meaning excavated but not to final grade. 122 There had been problems with leaks on the Friday before the incident but the major leak had occurred on Saturday, when Mr Mion had seen Dr Redman on site. There was an area on the southern wall where shotcrete had fallen off but Mr Mion thought that had not occurred on the Saturday. He could not remember whether there was another fall of shotecrete - his best recollection was that there was only one fall of shotcrete. He could not recall a conversation on the Saturday involving himself, Mr Di Cesare, Mr Shepherd and Dr Redman about the inability to install dowels in the shotcreting. 123 In relation to 21 October 1997, Mr Mion agreed that at the northern wall the bench was approximately 1 metre in width but he disagreed that it had been over-excavated. The 3.5 metre berm that was to be left was at the eastern wall and final bench width at the northern wall was meant to be half a metre. The leaks that had occurred on the northern wall and the eastern wall had brought sand in but all the sand had washed down to the eastern end. 124 On 21 October 1997, Mr Mion said he could see the width of the bench at the eastern wall although there was sand on top of it and mud and rubbish around it. It was not possible however to see the geometry of the bench at that time because of the surrounding material. 125 Mr Mion did not have a clear recollection of who spoke at the meeting with the men in the shed on 21 October 1997 and thought his statement of November 1997 might be more accurate than his present memory. He could not remember if Mr Di Cesare or Dr Redman spoke first. Mr Mion agreed that Dr Redman said that the need for temporary struts arose because of the shotcrete failure on Saturday and the water inflow. Mr Mion could not recall a number of matters put to him. He could not recall what was said by Dr Redman at the meeting. He agreed that Dr Redman said he was confident that the shotcrete problem had been addressed and the site was now safe. He did not remember Dr Redman saying that there was now a programme for the continued installation of anchors at the toe of the sheet piles at the eastern wall, although he recalled that anchors were being installed on the south side. It stuck in Mr Mion's mind that Dr Redman made some reference to the Class 2 or HAW2 sandstone and he doubted that he was mistaken in his recollection about that matter. He regarded Class 2 rock as hard rock. He felt it was safe to return to work because at least two people had said the cofferdam was safe. Mr Mion could not remember another meeting of the men taking place on the site on 21 October 1997 and he was not present at such a meeting. 126 On the morning of 22 October 1997, Mr Mion had used the 20 tonne excavator to level off the top of the bench in front of the eastern wall. It had been very untidy because there was a build-up of sand and slurry and the full width of the eastern wall was levelled. At the completion of the task, there would have been a flat area immediately in front of the sheet piles on the eastern wall extending from the northern point of the eastern wall to the southern point of the eastern wall. Mr Mion thought that the bench at the eastern wall after the levelling, had to be more than two metres wide but he could not be sure. On that day, there was no excavation occurring and by about 11.00 am all the leaks had stopped. Mr Mion recalled inspecting the wall at the eastern end with Mr Di Cesare but could not remember anyone else being present. He thought they discussed the next step to be taken. He recalled that there was drilling on the southern wall to install rock anchors but he did not think he had given any instructions to drill rock anchors in the eastern wall on 21 or 22 October 1997. Mr Mion said that on 21 October 1997 he did not recall a bucket from the excavator being placed against the eastern wall to stop the flow of water. The top of the bench was used to store items and different buckets could have been left there but there was no bucket used to stem the flow of water from a leak at the eastern wall. 127 By 22 October 1997, Mr Bahamondez had been employed for three years by Transfield Constructions Pty Ltd: on that date, he was working at the Cahill Park site. In an interview with Inspector Dubois, he explained how he had been swept away by the incoming water when the eastern wall collapsed and spoke of the injuries he received as a result. He talked about his ongoing disabilities. He had attended a meeting of workers a day or two before the collapse: sand had collapsed on the side of the dam and an engineer had attended but he did not know if that person was from Transfield Constructions Pty Ltd. 128 Mr Bahamondez gave oral evidence at the hearing with the assistance of an interpreter. At the meeting, Mr Di Cesare spoke first and the engineer after him. The engineer said that everything was safe and secure - everything was a 100 per cent safe. 129 In cross-examination, Mr Bahamondez agreed that there could have been two meeting of employees on the day before the dam collapsed. The first meeting took place in the crib office at morning tea and Mr Di Cesare spoke. He could not remember whether Mr Di Cesare was the only one to speak at the first meeting. He was able to understand what Mr Di Cesare was saying without the help of an interpreter when he referred to construction terms. There were people present who spoke Spanish and they would stand in a group and tell each other what was said. He could not recall to whom he had spoken in that group on the day of the meeting. 130 There was another engineer with Mr Di Cesare at the first meeting who was not from the company but Mr Bahamondez thought he was with the people who made the dam. That engineer was with Mr Di Cesare at the second meeting as well, but at the second meeting the other engineer spoke. It was at the second meeting that Mr Di Cesare said that the area was safe and that work could be performed there. At the second meeting held around lunchtime, Mr Di Cesare was the first person to speak. Shotcrete had fallen off the southern wall but he could not remember exactly what Mr Di Cesare had said. The only thing Mr Bahamondez could remember was that Mr Di Cesare spoke about the fallen shotcrete then spoke about security and that everything was safe. It was not only Mr Di Cesare who said that, it was the other person speaking as well - they were both saying it was all secure and safe. In the second meeting, they both spoke about the shotcrete falling. The second person who spoke was from the dam company. Mr Bahamondez was not sure that the second person who spoke at the afternoon meeting was also at the morning meeting: they were all office people. He could not remember whether anybody else spoke besides Mr Di Cesare at the morning meeting. On further examination, Mr Bahamondez said that the engineer who was at the first meeting with Mr Di Cesare spoke at the second meeting. Mr Bahamondez could not remember the second man referring to shotcreting or that there was a programme for installing rock anchors at the eastern wall. 131 In October 1997, Mr Worthy was employed by Transfield Constructions Pty Ltd as a leading hand/carpenter at the Cahill Park site and was also chairman of the Occupational Health and Safety Committee. He supplied a statement to Inspector Dubois concerning the steps he had taken following the collapse of the wall and the rescue of two workers at the site. 132 Mr Worthy said that, in the weeks prior to the wall collapsing, workers had been expressing concerns about the amount of water coming into the area. On Saturday 18 October 1997, at the southern wall, part of the shotcrete collapsed and concerns were raised. The wall was chipped away to inspect under the surface, the toe piles were exposed and Mr Worthy spoke to Mr Di Cesare who explained a Coffey Partners engineer's report to him. On 20 October 1997, several workers expressed concern to Mr Worthy about the amount of water coming into the excavation and cracks appearing in the shotcrete. Mr Worthy raised his concerns with his immediate supervisor Mr Patton, and was put in touch with Mr Di Cesare. Mr Worthy asked Mr Di Cesare to speak to the men to explain what was happening where the shotcrete was falling and the amount of water coming into the excavation. Mr Di Cesare said he would do so the next morning. At about 9.00 am on 21 October 1997, Mr Di Cesare spoke to workers in the cribroom. At approximately 10.40 am on the same day, at the southern wall under the surface area another part of the shotcrete collapsed. At 11.30 am, Mr Worthy spoke to Mr Di Cesare and informed him that the men were unhappy with what was going on. Mr Di Cesare said he would ask Coffey Partners to talk to the men. At 1.00 pm, an engineer from Coffey Partners spoke to the workers in the cribroom: that was Dr Redman. Dr Redman spoke to the workers in the cribroom and assurances were given that nothing in the area was structurally unsafe. 133 In oral evidence Mr Worthy said that the men had raised concerns about the safety at the southern wall where the shotcrete had collapsed. On 18 October 1997, he had raised the issue with Mr Di Cesare who told him that the area was basically safe. That response was reported to the workers. 134 Mr Worthy said there was concern about the amount of water coming in to the site, although it was normal, when working next to a river, to get some leakage. At the 9.00 am meeting on Tuesday 21 October 1997 in the cribroom, Mr Di Cesare told the men that the area in question had been checked by Coffey Partners and that the area was safe. After that meeting, the men were still concerned when shotcrete later fell from the southern wall and they continued to ask why the water was coming in. At the meeting in the cribroom, Mr Worthy said that Dr Redman had spoken about the southern wall, which was the problem area in question, and had told the men that it was safe enough to go back to work in that area. Mr Worthy thought that Dr Redman had addressed the meeting of men for about half an hour and dealt only with the southern wall. 135 Mr Kirkwood was employed as a carpenter by Transfield Constructions Pty Ltd at the Cahill Park site. He was present when the cofferdam failed on 22 October 1997 and participated in the rescue of two workers who were caught when the water swept into the site. He was present at two meetings before the wall failed where safety at the site was discussed. Mr Di Cesare had addressed the meeting and, at the second meeting, a consultant from Coffey Partners addressed the meeting. The engineers said there was no problem with water coming in and shotcrete falling off the wall: they had had a look and everything was allright as far as they could see. 136 On 21 and 22 October 1997, Mr Gray was working at the Cahill Park site as a contractor and was drilling holes for the insertion of reo bars and bolts. He had been directed where to place the bolts by the foreman Mr Mion. He commenced drilling on the northern wall and was working towards the eastern cofferdam wall. He firstly drilled eight holes, then moved to the southern wall. This was being performed on 21 October 1997 and, after a number of holes had been drilled, a large piece of shotcrete fell down into the excavation near where he was working. Mr Mion instructed him to drill into the area where the sheeting come away and he noted that, after four metres of drilling, the rock got harder and he drilled to six metres. Mr Gray continued with this drilling task on 22 October 1997. About 12.00 pm he had a few more holes to drill and was working approximately 40 feet from the eastern wall. He heard a noise and saw the bottom of the sheeting stick out on the eastern wall - water and mud began to gush in. He was caught up in the surge of water and was swept into the reo bars and other parts of the work and suffered a number of injuries which led to him being hospitalised. 137 After the accident, Mr Gray drew pictures of the state of the work as he recalled it immediately before the collapse of the eastern wall, and also the state of the works on 21 October 1997. Mr Gray said that he had not seen the excavator digging in the area to the east. At the eastern wall the spoil that was collected there was quite extensive and came out from the wall approximately 40 feet. 138 In cross-examination, Mr Gray said that, during the course of drilling the northern and southern wall, one rock bolt was put in on the southern wall but none on the northern wall. The rock bolts were being put in by someone else. Mr Mion's instructions were to drill only the northern and southern walls for rock bolts. Mr Gray had seen water running down the wall. On Wednesday, he did not see any additional water coming into the site before the collapse of the eastern wall. 139 On Wednesday 22 October, Mr Gray had not noticed that the spoil in front of the eastern wall had been levelled off nor did he see an excavator bucket or skip bin placed against the eastern wall. 140 During October 1997, Mr Lazar was employed by Transfield Constructions Pty Ltd as a rigger to work on the Cahill Park site. On the day before the eastern wall gave way Mr Lazar was working at the task of clearing the mud and slurry from the bottom on the excavation prior to the pouring of a concrete slab - this work continued on the day the wall collapsed. Mr Lazar witnessed the wall giving way and assisted in the care of two employees who were injured in the incident. As a result of the collapse of part of a wall there had been a meeting of employees on 21 October 1997 - that meeting was addressed by the engineers Mr Di Cesare and a person from Coffey Partners. 141 To the best of his recollection Mr Lazar said that the meeting of employees about work safety occurred in the cribroom just after lunchtime on the day before the collapse of the dam. He thought there were probably a dozen or more people there but he did not know the exact number. The engineer from Coffey Partners told the meeting it was safe to go down to the site, that it was just a piece of shotcrete that had fallen due to moisture in the wall, and, that the rest of the site was safe. The main concern of the men was working inside the dam when a piece of shotcrete fell down. The person from Coffey Partners did not speak about any other part of the site. 142 Mr Lazar had observed the excavator moving around the site cleaning it up. On 22 October 1997 the excavator was close to the sheet piles at the eastern end but not right against the pile - it was cleaning and scrapping the bottom of the excavation to make it clean. It was trimming, although Mr Lazar did not observe the excavator levelling the area immediately in front of the eastern wall on that day. He did not remember the excavator digging or levelling. 143 On the day before the eastern wall collapsed Mr Lazar attended one meeting of employees to discuss the safety of the site but he was unaware of any other meeting held that day. He agreed that at that meeting someone said words to the effect that they were confident that the shotcrete problem had been addressed and that the site was now safe. Mr Lazar also agreed that it was this statement that led him to say that he was assured that everything was alright - the shotcrete had been fixed and the site was safe. 144 On 22 October 1997 Mr Mauropoulis was continuing his employment as a crane driver with Transfield Constructions Pty Ltd and was working at the Cahill Park site. He witnessed the collapse of the eastern wall and assisted in the rescue of two workers who were swept away with the inrush of water. On the previous day he was in the crane when shotcrete fell off the southern wall. He said that everybody came out of the excavation during lunch and management spoke to the workers about the situation explaining that water had got behind the shotcrete and that it had just peeled off - it was a one off incident and the excavation was safe. He recalled Mr Di Cesare the ground engineer for Transfield Constructions Pty Ltd speaking in the lunch room about the excavation. 145 In oral evidence Mr Mauropoulis said that he could not remember precisely what was said at the meeting but the discussion lasted for about half an hour. The Transfield Constructions Pty Ltd engineer Max was present as well as the site manager Nino Bertoni. Max and the other engineer told the meeting that the excavation was safe and explained why the shotcrete had peeled away. The shotcrete fell from the northern wall because moisture had loosened the concrete. In cross-examination Mr Maurpoulis agreed that the other engineer who addressed the meeting said something about looking at all other shotcreted areas to establish there were no other circumstances where the issue of falling shotcrete could arise again. He agreed that he was anxious to hear that falling shotcrete would not occur in other areas of the excavation. 146 Mr MacDonald was the managing director of MacDonald Contractors. He gave a statement to Inspector Dubois concerning the contract his company had for drilling at the Cahill Park site. Mr MacDonald described this contract as "only a drill on dry hire". He had been told that Transfield Constructions Pty Ltd had the bolts on site and the order made no mention of providing anchors and grouting - the order was to drill holes as where directed by Transfield Constructions Pty Ltd. He had arranged for Mr Gray to attend the site and commence drilling while on the following day he received a message that Transfield Constructions Pty Ltd wanted to grout three bolts so he arranged for a crew to go to the site. The accident happened before that crew could enter the site. Mr MacDonald's recollection was that the job was "a bit of an emergency" to drill for the placing of rock bolts. Dr Redman's evidence about the Connell Wagner design 147 During the course of the hearing, the defendant raised the adequacy of Connell Wagner's design of the cofferdam. Inspector Dubois had, in a general way, expressed the view that the design was inadequate but certain matters raised during the evidence of Mr Shirley prompted Dr Redman to revisit a computer programme known as WALLAP. 148 Dr Redman's review of this computer programme was the subject of a separate statement which was supported by a number of documents relied upon by him as describing in detail the way in which the programme operated and the steps he had taken in reviewing its operation. He described WALLAP as being a commercially available computer programme widely used in the design of retaining structures. From other proceedings, he had been able to obtain copies of Connell Wagner outputs from WALLAP analyses - documents which he had not seen before 22 October 1997. 149 For the defendant, it was submitted that the evidence went to causation and responsibility and was also relevant to any s 53 defence that the defendant might wish to rely upon. This evidence was highly technical and, essentially, the technical nature of it was not challenged by the prosecution. Dr Redman stated that, for the purposes of this exercise, it was necessary to commence with an understanding that the defendant had provided a cohesion parameter to Connell Wagner for the purposes of Connell Wagner's design of the cofferdam. The parameter supplied by the defendant involved an undrained cohesion for HAW2 of 500kPa and certain values which were said to be consistent with an undrained assumption. Mr Shirley's evidence was to the effect that, although the WALLOP analysis showed the use of a drained cohesion parameter, that was due to a quirk of the computer system and was, in fact, the means by which an undrained cohesion parameter could be entered. 150 The purpose of Dr Redman's evidence was to test that evidence of Mr Shirley and then to see what result might flow if the programme was operated on the defendant's undrained cohesion for HAW2, rather than on what appeared to be Connell Wagner's use of a drained cohesion parameter. When the defendant's undrained cohesion parameter was entered, Dr Redman said that it showed a factor of safety of 1.928 when the excavation was at elevation 92. That was a safety factor which was acceptable. When the drained cohesion parameter for HAW2 was used, the factor of safety when the excavation was at elevation 92 was reduced to 0.984. Of this result, Dr Redman stated: This is an unsatisfactory design factor of safety, as failure is indicated. 151 Dr Redman also pointed out other features of the computer programme which drew attention to and warned against the use of incompatible values. The effect of Connell Wagner's failure to use the undrained cohesion parameter supplied by the defendant was that the design of the cofferdam was fundamentally flawed and that it was, indeed, designed for failure. There was some other evidence touching on this matter. Dr Gomes from Connell Wagner could not say what the significance was of using a drained cohesion, while Dr Redman stated that the use of a drained cohesion would be significantly over-estimating the strength of the material. DELIBERATION (a) Dr Redman's assurances of site safety 152 An issue which dogged the proceedings but, by consent of the parties, was left to be dealt with in final submissions was the prosecution's allegation that Dr Redman had spoken to the workers in the sheds on 21 October and had assured them that the site was safe when he had failed to make an appropriate assessment of the stability of the sandstone in which the sheet pile was embedded. The defendant firstly denied that was the effect of what was said by Dr Redman at the meeting of workers but, secondly, at an early stage, challenged the prosecution's submission that this matter was covered by the Particulars of the charge. 153 The two Particulars which highlight the nature of the breach were as follows: (c) The defendant failed to adequately assess the strength and integrity of the sandstone bedrock at the base of the cofferdam wall. (d) The defendant failed to assess the capacity of the sandstone bedrock to provide support at the toe of the cofferdam wall." In my view, it is significant that those two Particulars were then linked with the following Particulars, which stated: (e) As a result of the said failures, Juan Bahamondez and Darryl Gray suffered multiple injuries. (f) As a result of the said failures, a number of workmen were placed at risk of injury. 154 There is some significance for the prosecution case when the Particulars are framed in this fashion. The failure to adequately assess the strength and integrity of the sandstone at the base of the cofferdam and the failure to assess the capacity of the sandstone to provide support at the toe of the cofferdam wall resulted in two workers suffering multiple injuries and a number of workmen being placed at risk of injury. As so pleaded, the prosecution is required to demonstrate a causal nexus between the alleged failures and the risks to workers identified. In this case, the prosecution essentially looks back from the failure of the cofferdam wall, identifying that as the risk to workers: on the prosecution's approach it must follow that the failure of the cofferdam wall was linked to the failure of the defendant to adequately assess the strength and integrity of the sandstone at the base of the cofferdam wall and the failure to assess the capacity of the sandstone to provide support at the toe of the cofferdam wall. From early on in the prosecution case, however, it was accepted that the prosecutor could not demonstrate the reason for the failure of the wall but asserted that it did not need to do so. A part of the wall clearly came away, allowing an inrush of water which flooded the worksite. The sandstone in which the sheet piles were embedded was, at this point, very weak, indeed, so weak that, after the incident, Inspector Vierow, using diving equipment, was able to easily push a knife into the sandstone. 155 I find real difficulty in accepting the prosecution submission that Dr Redman's address to the workers on 21 October, assuming that assurances of safety were given, comes within the Particulars. While in one sense it might be said that the failure to make the assessments alleged meant that Dr Redman could not have assured the workers as to the safety of the site and allayed their fears, the address itself and its content is such a different subject as to warrant separate and precise particularisation. The defendant, reading Particulars (c) and (d) and linking them with Particulars (e) and (f), could not understand that the charge encompassed the content of Dr Redman's address to the workers in the sheds. The prosecutor asserted that the defendant was on notice of that part of the charge because there was a reference to the meeting in the sheds in the affidavit supporting the Application for Summons sworn by Inspector Dubois. In submissions, counsel for the prosecutor also asserted that the meetings of 21 October were evidence of the failure to adequately assess the strength and integrity of the sandstone. The defendant pointed out that the inspector's affidavit covered a wide variety of matters over a number of days before the period covered by the charge, and dealt with other aspects of the site arising at the northern and southern walls having nothing to do with the eastern wall where the collapse occurred. 156 While I entertain real doubt about Dr Redman's address as falling within Particulars (c) and (d), in any event, on the evidence, such particulars, if they formed part of the charge, are not made out. The evidence from those who attended the meeting in the sheds was not consistent and varied in a material respect, namely, whether Dr Redman was addressing the men about shotcrete falling elsewhere on the site rather than addressing the integrity of the eastern wall. He was not addressing the overall safety of the design or addressing the overall safety of the site. Dr Redman's evidence was that all the problems that had arisen over the weekend and which resulted in struts being inserted following the falling of shotcrete and the number of leaks were addressed by him and he told the workers that the steps taken had addressed those issues. He did mention, but it appears to have been a minor part of his address, that rock bolts would be inserted at the eastern wall. Dr Redman stated that he was primarily addressing the falling shotcrete and the leaks occurring elsewhere. A number of employees who attended the meeting addressed by Dr Redman gave evidence and their evidence has been summarised. There is no suggestion that these employees took a contemporaneous note of what was said at the meeting: they were relying on their memories of an event now long past. It is true that a number of employees thought that, at the meeting, they had been told that the site was safe, but they were far from unanimous about who had given that assurance and, looking at the evidence in its totality, it appeared to be more an impression gained from what was said rather than a precise statement made by anybody and, in particular, made by Dr Redman. A number of employees who attended the meeting and, notably, the chairman of the safety committee thought that only the southern wall was being referred to and the problem of falling shotcrete. 157 I accept Dr Redman's evidence that he was addressing a much narrower issue directly arising from matters with which he was recently involved, namely, the leaks that had occurred over the weekend, the falling shotcrete and the installation of the strut to give support to the dam. Connell Wagner, as the prosecution accepts, was the designer of the dam and the defendant was engaged to assist in that task by providing geotechnical advice. It was consistent with that role that Dr Redman addressed those matters with which he was directly involved rather than the overall design of the dam of which he knew little until well after this incident. The prosecution, on this evidence, was unable to make out its case, even if this meeting and Dr Redman's address could be said to fall within the existing Particulars. (b) What were the defendant's contractual obligations and what tasks were undertaken? 158 In opening, the prosecutor stated that it was unlikely that the true nature of the contractual arrangements between the parties would be evident on the evidence. The evidence did establish that the defendant was to supply geotechnical parameters to Connell Wagner relating to the construction of the cofferdam by Transfield Constructions Pty Ltd. Transfield Constructions Pty Ltd was building the railway and Connell Wagner was designing the cofferdam and, in the course of that task, from time to time sought geotechnical advice from the defendant. The defendant repeatedly made the point that it was not the "designer" of the cofferdam, a point accepted by the prosecutor, who nevertheless submitted that the provision of geotechnical advice involved the defendant in the design of the cofferdam. Connell Wagner also certified that the works were being constructed in accordance with the design. At an early stage, the defendant offered to assist in the design of the cofferdam but that did not occur. Indeed, during the tender stage, the defendant had warned Connell Wagner that the design of sheet pile walling was geotechnically complex, which led the defendant to offer assistance in the design of the cofferdam. Dr Redman had not seen design calculations performed by Connell Wagner for the project and the defendant had never been engaged for the detailed design phase of the cofferdam project. In the absence of a design review the defendant did not see its role as "second guessing" the designer, especially after advising Connell Wagner on 3 October 1997 to avoid excavating rock near the sheet piles until they were confident that the toe support was appropriate for the loads being experienced. 159 Dr Redman's evidence was that the role of the defendant in design did not extend beyond comment at the concept design stage, and that the defendant provided input to Connell Wagner regarding the geotechnical model and parameters. This involved the defendant from time to time carrying out onsite visits and providing advice as requested. The defendant was not provided with final drawings showing the support designed for the eastern wall nor was the defendant asked to provide comment on any WALLAP models or other design calculations for the cofferdam design. The defendant asserted and there was no evidence to the contrary that only the designer fully knew the assumptions relied upon in its design and therefore, what aspects needed inspection or verification during construction. 160 The nature of the geotechnical investigations carried out by the defendant in Cahill Park are set out in par [66] et seq. The defendant geotechnically modelled the subsurface conditions for the entirety of the railway line. The defendant's investigations involved a number of regional studies of bedrock elevation, including the region between Redfern and Wolli Creek. There were 12 boreholes in the 300m section at and about the location of the failed eastern wall. Using a well-understood system of classification of sandstone, the defendant introduced its own classification of the sandstone for the purpose of this project. That classification system encompassed the traditional system of classification, but in broader strata: having indicated that the sandstone would vary in its strength, the defendant concluded that the sandstone in and around the cofferdam would be HAW2, that classification encompassed moderately weathered sandstone. Under this system, HAW3 would represent severely weathered sandstone unable to provide real support for sheet piling in a cofferdam. The defendant further supplied a cohesion parameter of 500kPa for this moderately weathered sandstone. 161 From the variety of expert evidence that was called, including that of Dr Redman, there is no evidence that calls into question the geotechnical modelling performed by the defendant nor its conclusion that at the cofferdam the sandstone would fall into the category of HAW2. There was a period when the prosecutor alleged that, at the eastern wall of the cofferdam, the sandstone was very weak and was, in fact, HAW3 and not HAW2 as assessed by the defendant. Professor Carter took borehole samples of rock from around the eastern wall and, although their strength was low, they nevertheless fell just within the minimum level which would constitute the rock as HAW2 under the defendant's system. Ultimately, the prosecution desisted with the allegation that the sandstone at the site of the failed wall was HAW3, rather than HAW2. 162 There was no suggestion in Professor Carter's evidence, or in the evidence of any other expect, that there was any error in the geotechnical modelling performed by the defendant either generally or in a way which affected the stability of the eastern wall of the cofferdam. Professor Carter took a number of borehole samples of the rock, but there is no suggestion that they demonstrated a different strength of rock to that assessed by the defendant. Nor is there any evidence that the boreholes were not sufficient in number or appropriately located to make an appropriate assessment of the strength of the sandstone at the cofferdam. As earlier recorded, it was said by the prosecutor that it was no part of its case that Dr Redman was slapdash or cavalier, nor was Dr Redman required to demand Connell Wagner provide design parameters to the defendant to enable it to assess the adequacy of the design. Dr Redman was not being asked to advise whether a cofferdam built to Connell Wagner's design would collapse. Rather, the allegation was that the defendant had failed to factor in a number of possibilities which ultimately led to the failure of the cofferdam wall. In short, it was not the prosecution case that anything done by the defendant in the exercise of its professional role of providing geotechnical advice and modelling amounted to a failure to identify a critical fact or factors which would lead to the collapse of the eastern wall of the cofferdam. The prosecution case then comes down to the errors or omissions allegedly made by the defendant in its inspections of the site at various stages of construction and the failure in those circumstances to assess the strength of the sandstone at the eastern wall and at the foot of the eastern wall of the cofferdam on 21 and 22 October 1997. (c) A failure to assess the strength of the rock? 163 In opening, the prosecutor summarised, in ten steps, how the alleged failures of the defendant arose: 1. a dangerous situation had clearly arisen in the site from at least the Saturday prior to the collapse of the eastern wall; 2. experienced workers had drawn their own conclusions and had demanded safety assurances; 3. a meeting was called for an expert being an engineer and ultimately Dr Redman as an appropriate expert to discuss with the workers their concerns; 4. the meeting addressed by Dr Redman took place at around lunchtime on Tuesday, 21 October 1997; 5. at that meeting, Dr Redman on behalf of the defendant, assured the workers and their supervisors including Transfield Constructions Pty Ltd that the site was safe; 6. relying on those assurances, work continued at the site on that day and the following day; 7. the site was in fact extremely unsafe on those two dates as ultimately evidenced by the dam's collapse on Wednesday, barely 24 hours after the site meeting in which the assurances had been given by Dr Redman as to the site's safety; 8. that the collapse took place largely because of inadequate strength of the cofferdam structure at the base or toe of the eastern wall 9. that the inadequate strength was because of a combination of factors, including possibly, on the evidence, inadequate embedment depth of the sheetpiles into the sandstone bedrock, inadequate berm width at the rock of the toe of the eastern sheet piles and inadequate remedial measures taken for this situation, including there were no toe anchors actually put into the eastern wall before the collapse - whatever caused the collapse it was because of factors relevant to the anchoring of it at its base or toe or foot by the bedrock sandstone and factors relevant to the strength of the sheet piles themselves or the way they were interlocked; 10. if the defendant had adequately assessed strength and integrity or assessed the capacity of the sandstone at the base or toe of the cofferdam wall the defendant would have been aware of the inadequate strength of that support and Dr Redman could not have given the assurances he did to the workers on 21 October 1997 which directly led to their being at risk. 164 As raised with counsel for the prosecutor during addresses, the manner in which the charge is framed relies on a failure to assess the strength and integrity of the sandstone rock at the base of the cofferdam and to assess its capacity to provide support at the toe of the cofferdam wall. The evidence shows numerous instances of the defendant inspecting and/or assessing the sandstone rock. This is not a case where there was a total failure to carry out inspections or a total failure to make assessments of the stability of the rock but, as will be disclosed later, the prosecutor's case became much more confined as it was progressively required to discard generalities and focus on demonstrated acts or omissions having regard to the totality of the evidence. This refinement by the prosecution led to continuing complaints from the defendant that the allegations now made were quite different from those particularised. There was more than some substance to that complaint. 165 It should be clarified at this point that the evidence did not demonstrate any continuing contractual obligation on the defendant to attend and inspect the berm in front of the cofferdam walls or any other part of the site. It was contemplated that, as the excavation reached a level where the berm would be trimmed back to its final width, the defendant could be called upon to make assessments with Connell Wagner as to the capacity of the wall to withstand the pressures of the outside river water. In a document provided by Connell Wagner to Transfield Constructions Pty Ltd dated 30 June 1997 addressing the issue of toe support being the quality of the HAW2 rock, it was stated that lack of toe fixity may occur if the pile was driven into very jointed HAW2 or if the driving crushed the HAW2. It was said that it was known that the pile was embedded the required distance into HAW2 for toe fixity but, if the quality of HAW2 was poor, it was unlikely that any effect would be noticed on pile fixity until the rock excavation began near the pile toe. Although it was thought unlikely to be necessary, a methodology was proposed whereby the lack of pile fixity would be monitored as excavation proceeded down to the rock and lack of pile fixity flagged from observing the load/deflection characteristics of the wall. A procedure was recommended that involved monitoring pile movements and strut forces using compression load cells. They would be monitored continually for any sudden change. If excavation proceeded towards the base without any sudden changes in load/deflection then potential jointing of the rock or poor quality rock was to be identified and appropriate remedial action taken. It was proposed that, upon reaching 2.5 metres of the pile base, a slot would be excavated at right angles to the wall and in the centre of the cofferdam, and Connell Wagner and/or Coffey Bros would review the depth and quality of the rock to determine if toe support was required. If it was required, then a rock anchor design was referred to. The document then refers to the position if the vertical face below the bench was excavated, then regular inspections by Connell Wagner/Coffey Bros would determine the need for additional rock bolting to support against wedge failure. 166 The significance of this document, amongst other things, is, firstly, that it proceeds on the basis that either Connell Wagner or the defendant, or both, would conduct the assessments of the cofferdam walls and, secondly, would be involved in the decision to support the walls, if necessary, by rock bolting. Dr Redman's evidence was that he did not see this document until after the collapse of the wall and it would appear that the expectation that the defendant would attend from time to time to carry out inspections was not reduced to a continuing contractual obligation but arose, as described by Dr Redman, on an "as required" basis. 167 Two other factors regarding this project added to the difficulty of the case particularised by the prosecutor. Firstly, it was accepted that this was a "design and construct" project and that, during the construction phase, a number of proposals would be considered, including the installation of rock bolts and discussions as to the dimension of the berm to be left in front of the sheet piles. The project manager for Transfield Constructions Pty Ltd, Mr Shepherd, described a "design and construct" project as one where the client gave conceptual design and design parameters to the contractor, to consider them and then design and construct the structure, including the detailed design, gathering all the design inputs and producing a set of design outputs or deliverables which were handed to the client for review. At this site, the design was occurring, consistent with a design and construct contract, throughout the course of the works as circumstances required. It also involved an observation method whereby Connell Wagner an/or the defendant, once the site had progressed to certain levels in the excavation, would excavate test pits and the geological model or assumptions made would be verified. The geological model was based on interpretation of the ground between boreholes, acccepting that it was in the nature of sandstone rock to vary in strength. Connell Wagner had proceeded on the basis that there would be no "toe fixity", that is, that there would be poor quality rock at the base of the sheet pile which would not provide lateral restraint. On this assumption, an inspection stage was required to assess the quality of the rock as excavation approached the toe of the sheet pile to assess the strength of the rock. This might involve the sinking of boreholes, but however the assessment was made, if the rock was of inadequate strength, then rock bolting would be used to provide the necessary support to the wall. 168 The design and construct approach, together with the no toe fixity approach, caused difficulties for the prosecution as much of its case appeared to rely upon a fixed approach to design and a failure to come to grips with the no toe fixity approach of Connell Wagner. These two elements of the project at Cahill Park meant that there would be ongoing issues as the site developed which would require attention according to the problems exposed. In this context, a prosecution case framed in simple terms and based upon a failure to assess rock strength around the base of the eastern wall does not adequately comprehend or reflect the nature of the tasks being undertaken and, in particular, undertaken by the defendant. 169 The critical issue then becomes the acts or omissions of Dr Redman on behalf of the defendant on the day before the wall collapsed. Although the charge revolves around 21 and 22 October 1997, the prosecution put the charge in the context of events occurring since 18 October, with the site experiencing leaks and the falling of significant amounts of shotcrete. Over the weekend, with the assistance of advice from the defendant, supporting struts had been erected but this combination of events had caused concern amongst the workforce. It was this combination, the prosecution asserted, that should have signified to Dr Redman that the cofferdam was under such pressure that urgent remedial action was required, including at the eastern wall. 170 Dr Redman had considered these events and, by 21 October, had determined that there was no point in going through the exercise of sinking boreholes to examine the nature and strength of the rock near the eastern wall because he had already formed the view from what he had seen elsewhere on the site that he should go directly to having rock bolts inserted to support the eastern wall. The prosecution case asserted that Dr Redman should have taken the step of actually sinking the boreholes because that step would have demonstrated to him that the rock was very low in the HAW2 range and weathered to the point where it could not be relied upon to support the wall. The real difficulty with this approach is that, if Dr Redman had taken the time, whatever time was involved, to sink the boreholes and to obtain the samples that would demonstrate the weakness of the rock at the eastern wall, the remedial action was still to insert rock bolts to support the wall. If sinking the borehole or passing over that step still led to a decision to rock bolt the wall, what was the failure of assessment by Dr Redman and the defendant? The prosecution response appeared to be that if Dr Redman had followed the process of sinking the borehole and physically seeing the weakness of the rock, he would not only have recommended rock bolting but would have, firstly, understood that the wall was in imminent danger of collapse and, secondly, would have directed that the site be cleared until rock bolting or other support of the eastern wall had been completed. 171 There are insurmountable obstacles to the acceptance of this prosecution submission. It is well-established on the evidence that the strength of sandstone is variable and was variable within the site. There is no evidence that establishes that if one or more boreholes were sunk anywhere along the base of the eastern wall, the core retrieved would show only very weak HAW2. There is no evidence that a certain minimum number of boreholes needed to be sunk and that such a minimum number of boreholes would demonstrate the existence of very weak HAW2. There is no evidence from any expert that, if the borehole's core had shown very weak HAW2 in the immediate vicinity of the eastern wall, that a person of Dr Redman's qualifications and standing would have determined that the wall was in imminent danger of collapse such that it would be necessary to evacuate the site and take remedial measures. The prosecution assertion that the wall was in imminent danger of collapse does not arise from the evidence of experts but rather from the evidence that a day later, it did in fact collapse. Moreover, from the opening of the case, the prosecution accepted that it could not prove the precise reason for the collapse of the wall, and suggested it might lie in a combination of over-excavation of the berm, lack of embedment in the rock, the weakness of the rock and even, possibly, an aspect of the sheetpiling itself. With the prosecution case in that state of understandable uncertainty, the prosecution is quite unable to make out its case on the basis that observation of a rock sample taken from the base of the eastern wall would inevitably lead Dr Redman to the realisation that it was in imminent danger of failing. As the defendant put in submissions, the rock bolts might have been inserted yet the wall might still have failed: the evidence does not exclude that possibility. 172 The prosecution assertion that the circumstances of the previous few days and an inspection of a borehole core would have alerted Dr Redman and the defendant that the eastern wall was in imminent danger of collapse requires some consideration of those surrounding circumstances. To the extent that those circumstances include movement in the cofferdam walls, falling of large sheets of shotcrete, the existence of leaks at various parts of the dam and the insertion of supporting struts, there is no expert evidence that directly links those events with the failure of the eastern wall. Indeed, at one point counsel for the prosecutor submitted that the case revolved around the inherent weakness of the sandstone at the base of the eastern wall, which was probably HAW3 although there was no evidence to that effect, and that Dr Redman should have known of this defect in the strength of that rock. These submissions give an indication of the prosecution's shifting case even at the point of submissions, a matter which again drew objection from the defendant. The evidence shows that the defendant was not aware that Connell Wagner had decided to use load cells to monitor the pressure on the cofferdam walls and neither Connell Wagner or Transfield had informed the defendant of load readings. Despite a number of expert witnesses being called, no-one made the direct link between these factors and the failure of the eastern wall on 22 October 1997. The prosecution submissions in that regard are without support. 173 In considering the adequacy of the assessments made by the defendant in relation to the eastern wall, it is significant that other experts did not criticise the methods and parameters put forward by the defendant. Dr Gomes stated that Connell Wagner had advised a berm of 3.5 metres as being sufficient support to enable inspection before further excavation occurred and allowed inspection to determine the quality of the sandstone. Depending on the quality of the sandstone, rock bolts could be used. This is consistent with the course adopted by the defendant. 174 Dr Gomes stated that the defendant had conducted extensive investigations of the rock through which the project was to pass and had excavated a number of boreholes. There was at least one borehole which had been excavated within close proximity of the eastern wall of the cofferdam. This evidence suggests that the defendant had a core sample from rock close to the eastern wall but there is no suggestion amongst the experts that the strength of this rock was inadequately assessed by the defendant. Dr Gomes understood the classification HAW2 adopted by the defendant and also understood that this class of sandstone was quite variable in nature. Allowing for this variability of sandstone, there was no criticism by Dr Gomes of the assessments made by the defendant. In addition, Dr Redman said that there would always be uncertainty as to the rock strength regardless of the number of boreholes excavated - this evidence was not challenged. 175 Dr Gomes offered the opinion that the observational technique used on this site required fairly prompt action in the event of circumstances arising such as the requirement for rock bolting. When Dr Redman made the assessment on 21 October 1997 that there should be no further excavation and rock bolting should take place at the eastern wall because toe support could not be assumed, that opinion clearly constitutes the type of circumstance referred to by Dr Gomes where prompt action was required. Mr MacDonald from the drilling company regarded his engagement as meeting "a bit of an emergency". It is apparent from this evidence that Connell Wagner were alert to the necessity for prompt action once Dr Redman's opinion was made known - this evidence hardly supports the prosecutor's criticism of the lack of urgency surrounding this advice. 176 In relation to the evidence of Professor Carter, he does not appear to have been asked to look at core samples taken by the defendant to determine the accuracy of the defendant's assessment of them, and he says nothing about the appropriateness or inappropriateness of the bore holes excavated by the defendant for the purpose of assessing the strength of the sandstone at the eastern wall. There is simply no attack on the adequacy of the defendant's assessments of the rock at the site and at the eastern wall. 177 Mr Shirley, who provided a report to the WorkCover Authority of New South Wales, noted that the bed rock in the vicinity of the excavation varied significantly in quality and type but was typical of the upper layers of bed rock encountered in excavations within Hawkesbury sandstone. This is consistent with the classification of HAW2 adopted by the defendant and the evidence of Dr Redman as to the variability of sandstone to be found at this worksite. 178 Mr Shirley stated that the geotechnical parameters provided by the defendant were reasonably compatible with the local experience of Hawkesbury sandstone. He believed that toe anchors should have been used before the width of the berm had been reduced and excavation had continued but this was the approach adopted by Dr Redman for the defendant. The difficulty was that Dr Redman was unaware that the berm at the eastern wall had been reduced to 2 or perhaps 2.5 metres when it should have been 3 to 4 metres in accordance with his advice. That advice was consistent with Mr Shirley's opinion that it was always necessary to have a substantial berm in front of the wall before installing rock anchors. He also agreed to the berm of 3.5 metres or even 4 metres would mean that the wall was unlikely to fail. 179 There were some reservations expressed in Mr Shirley's report because of the lack of opportunity to fully discuss all aspects of design and parameters with those involved including the defendant, and in evidence Mr Shirley stated that he was not aware that Connell Wagner had modelled the cofferdam on a no toe fixity approach. It was his view that the sheet piling was under designed, having regard to the parameters supplied and that the design was grossly over stressed. This conclusion, even allowing for the reservations expressed, does not amount to a criticism of the assessments made by the defendant. The prosecutor, significantly, relied on Mr Shirley's evidence that a wedge of sandstone had "popped out" - it was not that the sandstone had completely collapsed. This evidence tends to undermine the central aspects of the prosecution case rather than support it. 180 The prosecutor's attack on the defendant for its alleged failure to properly assess the strength and quality of the rock at the toe of the eastern wall seems to overlook the fact that Connell Wagner, the designer of the cofferdam, was proceeding on a no toe fixity approach. According to Dr Redman's evidence, this approach assumed that any contribution being made to the toe from the HAW2 rock was less than had been recommended and this was reflected in the document sent by Connell Wagner to Transfield Constructions Pty Ltd in 1997. That document also factored in the possibility of the presence of blocks or wedges or jointed sandstone. Dr Redman was not sure whether this document assumed that there would be no support whatsoever in view of the fact that a temporary berm would be providing some restraint, but the no toe fixity scenario was essentially one based on very poor rock quality with the designer appearing to assume that there would not be any significant lateral restraint at the toe. The point of the inspection stage was that was the time a decision would be made as to the strength of the rock and, if it was assessed to be of insufficient strength to give support to the wall at that level of excavation, then rock bolts would be installed. This is precisely the step that Dr Redman took on 21 October 1997 when he advised that rock bolts should be installed, that there should be no further excavation until that task had been completed and that no support should be assumed at the toe. I accept the defendant's submission that the no toe fixity approach operated quite independently and regardless of the strength of the rock. Inspector Dubois accepted the same point as being accurate. Further, Dr Redman asserted without contradiction that the defendant did not determine the toe fixity. The evidence simply does not establish a failure by the defendant to assess the strength of the rock as particularised. Dr Redman made many assessments and none were called into question by any expert. 181 In relation to the berm width, the prosecution case does not allege that the various assessments made by the defendant and advice that there should be a 3 to 4 metre berm left in front of the cofferdam wall up to inspection stage was wrong or in any way relevantly inaccurate. The expert evidence, apart from Dr Redman, accepts that there needed to be a sizeable berm and within the range specified by the defendant. The prosecution attack focuses on the fact that the defendant did not inspect the berm at the eastern wall on 21 or 22 October 1997 because inspection would have revealed that the berm had been over-excavated probably to 2 metres. It is pointed out that, on 18 October 1997, Dr Redman, while on an inspection of the site, walked past the eastern wall and could have easily inspected the berm at that time although 18 October 1997 is not a time particularised by the prosecutor when a breach of the Act was committed by the defendant. Prior to 21 and 22 October 1997, the defendant through Dr Redman had advised of the necessity to keep a substantial berm between 3 and 4 metres and had, in September 1997, criticised Transfield Constructions Pty Ltd for over-excavation and had ordered backfilling as a remedial measure. Dr Redman regarded the advice given at that time as clearly indicating to Connell Wagner and Transfield Constructions Pty Ltd the necessity of maintaining a 3 to 4 metre berm and the dangers of over-excavation. There appears to have been no contractual obligation for the defendant to continue inspections and they could have been just as effectively carried out by Connell Wagner using the parameters supplied by the geotechnical specialist, the defendant. 182 Dr Redman's evidence is that on 18 October 1997, he had been told that there was a leak at the eastern wall but that it was being attended to by a grouting company as was the usual practice. The leak had been described as "minor" on this day. He had not been alerted to any concern by the grouting company and, during the course of inspection of the site, had walked past the eastern wall. At that time, there was considerable spoil over and around the berm such as to hide the actual dimensions of the berm. The evidence makes it difficult to determine precisely when the berm was over-excavated and trimmed to approximately 2 metres. This appears to have been the situation on 22 October 1997 when Dr Redman was not at the site and there is some unsatisfactory evidence that some trimming may have occurred on that day, but that evidence is not firm or conclusive. The size of the berm on 18 October 1997 is unknown and, considering the common evidence that there was a large amount of spoil around the eastern wall at that stage, it might be thought unlikely that the over-excavation or trimming had occurred by that date, although it is possible that it had occurred and that the spoil was deposited after over-excavation had taken place. The evidence presents three possibilities as to when the berm at the eastern wall was reduced to 2 metres: firstly, in the month before the failure of the wall; secondly, about 18 October 1997; and, thirdly over-excavation on the day of the collapse. The evidence is contradictory and no single possibility presents as being the most likely time as when the berm was reduced. With the evidence in this state, it is quite impossible for the prosecutor to establish that on 21 October 1997 the defendant should have known that the berm had been dangerously reduced - the evidence still leaves a possibility that this did not happen until 22 October 1997 when the defendant was not on site or required to be on site. 183 In dealing with this allegation, it is to be remembered that the defendant, on the prosecution case, was providing advice on an ad hoc or as required basis. There was no particular engagement of the defendant to inspect the berm at the eastern wall on 21 or 22 October 1997 and, in any event, the defendant had advised Connell Wagner and Transfield Constructions Pty Ltd to maintain a berm which all the experts accept should have supplied sufficient support for the wall. How then did the defendant's obligation arise to conduct an inspection of the eastern wall berm on 21 and 22 October 1997? The best that the prosecution case could assert was that previous over-excavation by Transfield Constructions Pty Ltd should have alerted the defendant that Transfield Constructions Pty Ltd was deficient in this area and its employees over-zealous. The evidence does not support a finding of predisposition to over-excavation on the part of Transfield Constructions Pty Ltd but, to the contrary, that over-excavation had been met with a very firm response from the defendant that the berm width had to be maintained. Dr Redman had seen nothing on site thereafter to suggest that this advice had not been followed. This evidence falls far short in establishing the criminal liability of the defendant. 184 While the evidence is unsatisfactory in pinpointing when the berm at the eastern wall was reduced from 3 - 4 metres to 2 - 2.5 metres, the evidence does suggest that, on the day of the collapse, there was cleaning work and trimming work being conducted at the eastern wall. It is not suggested that any excavation which took place around the berm of the eastern wall on 22 October 1997 should have been known by the defendant in circumstances where Dr Redman was not present on the site at that time. Although, on 21 October 1997, Dr Redman had decided that rock bolting was appropriate for the eastern wall, Mr Shepherd had accepted at that point that inspection of the eastern wall berm was not scheduled. There does not appear to be anything in the evidence which suggests that Dr Redman should have been aware of over-excavation of the eastern wall berm or that he was obliged to conduct an inspection of the eastern wall berm and failed to do so. 185 In addition to the above matters, I have had some difficulty in determining how a failure to inspect the berm, if made out, fell within Particulars (c) and (d) constructed as they are by reference to the strength and integrity of the sandstone bed rock at the base of the cofferdam wall and the toe support at the cofferdam wall. Particulars drawn in this manner appear to concentrate on the embedment of the sheet pile and the poor quality of the rock into which it was driven. That is how the prosecutor opened the case. The Particulars do not seem to address an inspection of the adequacy of the berm: that would require a very generous reading of the Particulars. In light of the view I have formed of this part of the prosecutor's case, I do not need to finally determine this issue. 186 A number of miscellaneous issues should be briefly dealt with: (a) on 3 October 1997, the defendant sent a site note to Connell Wagner expressing concern at support being provided for the cofferdam walls. In that document the defendant advised Connell Wagner to avoid excavation of the rock near sheetpiles until they were confident that the toe support was satisfactory for the loads being experienced. Connell Wagner did not advise of any problems or concerns following the receipt of this advice. That advice was followed by the further site note of 21 October 1997 advising the installation of rock bolts and warning that support at the toe of the dam should not be assumed. Connell Wagner and Transfield Constructions Pty Ltd were again warned that there should be no further excavation. The sequence of rock bolting was to be undertaken by Connell Wager indicating that the implementation of the advice was in the hands of the cofferdam designer and the constructor. These site notes also constituted appropriate assessments made by the defendant as to the strength of the rock and its ability to support the eastern wall; (b) the 21 October 1997 site visit by the defendant was made for the purpose of observing the conditions exposed in excavated faces near the service area and to arrive at recommendations for any further shotcreting, lagging or other support to the exposed faces. This was clearly set out in the site note of the same date and tends to support the conclusion that this site inspection was not as a result of any arrangement or contractual obligation on the defendant to inspect the berm at the eastern wall nor to inspect the rock at the foot of the eastern wall which had not yet been exposed; (c) the defendant's site note to Connell Wagner of 21 October 1997 dealing with the eastern wall spoke of consideration being given to the overall sequence of matters to be attended to in order to "ensure there are no delays to anchor bolt installation or shotcreting ...". This part of the site note confirms the immediate importance of the steps advised but does so on the understanding that it is Connell Wagner and Transfield Constructions Pty Ltd that were to take those steps; (d) the evidence established that a new wall was constructed 5 metres further away from the failed wall and that the same construction approach was taken to complete the task. This evidence supports a conclusion that there was no error in the defendant's geotechnical modelling and general assessment of the strength of the rock, accepting as the defendant did, that the rock strength would be variable; (e) the prosecutor relied on the presence of leaks in the cofferdam over the weekend before the collapse of the eastern wall, including evidence of a leak at the eastern wall as part of the factors that the defendant should have taken into account and which should have led to a conclusion that the eastern wall was in imminent danger of collapsing. During the course of addresses the prosecutor accepted that every leak could not be taken as evidence of a risk of imminent failure of the cofferdam. The evidence showed that leaks in a structure such as a cofferdam were not uncommon and were usually addressed by grouting and plating. There was a suggestion that the leak at the eastern wall was much more serious although when Dr Redman was first called to the site in relation to the concreting, no mention was made of the eastern wall until he raised it and was told that there was a minor leak at that time. There was no suggestion to Dr Redman that it had turned into something major and he was never told that the leak was "surcharging". Further, the evidence suggesting that the leak was so bad at the eastern wall that the excavator driver placed a bucket over it to stem the flow of water was in fact rejected by the excavator driver in his evidence. All of these matters detract from the prosecution case. (d) An alternative hypothesis 187 In meeting this case, the defendant relies on two further matters. Firstly, it asserts that on Dr Redman's evidence the design or the application of it was seriously flawed in that Connell Wagner had failed to properly use the cohesion parameters provided by the defendant, with the consequence that the predicted factor of safety was overstated. Secondly, it is argued that unknown to the defendant the berm at the eastern wall was over-excavated and below a level which would adequately support the sheet piling. The evidence as to the proper use of the cohesion parameter provided by the defendant was solely that of Dr Redman - and no other expert entered this field or was called to deal with the issue in detail. Dr Redman's eminence in this specialist field was accepted by other expert witnesses and the Court has no basis to doubt his conclusion that the misapplication of the cohesion parameter provided meant that the cofferdam was bound to fail. In relation to the over-excavation of the berm, the prosecution case was not that the defendant should somehow have known that fact but rather that the defendant should have conducted an inspection of the berm when the over-excavation would have been disclosed. In relation to both these matters, it may well be that these two issues represent an acceptable hypothesis which is reasonably consistent with the innocence of the defendant, leading to the conclusion that the case against the defendant has not been proved beyond reasonable doubt. On the view that I have taken of the evidence, the prosecution has failed to make out the case it has particularised and it is unnecessary to finally determine the alternate hypothesis developed by the defendant. 188 In all these circumstances the case against the defendant cannot be sustained. The order of the Court is that the summons be dismissed. GLOSSARY bench Mass of rock in front of sheetpile toe, also known as the berm. berm Mass of rock in front of sheetpile toe, also known as the bench. box pile Type of sheetpile with hexagonal cross-section. cohesive strength Also known as sheer strength, lateral toe restraints, toe. See toe restraint. declutch Situation where the joints between adjacent sheetpiles loosen, after permitting water ingress. dowel Long, relatively narrow cylindrical metal rod. Normally threaded. Forms part of anchor/rock, anchor/rock bolt/toe anchor. east wall The eastern wall of the sheetpile cofferdam being box piles and sheetpiles numbers P17 to P31. HAW2 A classification of sandstone, combining Classes 111 and IV from Pels (1978) classification scheme. HAW3 A classification of sandstone, equating to Class V from the Pels (1978) classification scheme. lagging Process of welding horizontal sheetpiles under each other. Used by Transfield to retain water-charged sediments. Used between the usual vertical sheetpiles: for example, under water main on North and South walls of cofferdam where vertical sheetpiles could not be driven. lateral toe restraint Also known as cohesive strength, sheer strength, toe restraint. See toe restraint. no toe fixity scenario The scenario where the toe of the sheetpile would not be laterally restrained by the material into which it was driven. That is, the scenario where the toe of the sheetpile was driven into very weak rock. Pels (et al) A classification system developed by the Australian Geomechanics Society in 1978. The system classifies sandstone and shale rock of various qualities from Class 1 (highest quality) to Class V (lowest quality. Princess Highway Crossing site That part of the proposed route for the New South Railway bounded by chainages 9240 to 9399 (ie approximately 159 metres long) located in Arncliffe, NSW. RL Reduced Level24 in metres. The mean seal level was RL 10025 while the sheetpile toes on the eastern wall were approx. at about RL 90. rock anchors Method of restraint/strengthening whereby dowels inserted into rock and grouted with external end restrained. rock bolts See Rock anchors. shear strength Also known as cohesive strength, lateral toe restraint. See toe restraint. sheetpile Steelpile (generally) driven vertically down into rock. Two different cross sections: (1) essentially flat; (2) hexagonal (called "box" pile). Interlinked via clutches. shotcrete Concrete sprayed onto vertical faces of rock, usually on to mesh anchored into rock face with dowels. Used to provide protection to rock face. toe anchors Anchors drilled through sheetpile to provide toe restraint. toe fixity The concept of the toe of the sheetpile toe being laterally restrained. toe restraint Normally used in terms of the lateral toe restraint provided by rock at toe of sheetpile(s): ability of rock at sheetpile toe to resist the force of the water on sheetpile (ie ability to resist the load). Also known as cohesive strength, restraint, sheer strength. See toe restraint. Waler Horizontal structural member attached along/between sheetpiles. oo00oo Bertolli Michael Antonelli, a project engineer with Transfield. Coffey Coffey Partners International Pty Ltd ACN 014 550 277, the defendant. Connell Wagner Connell Wagner Pty Ltd ACN 005 139 873. Di Cesare Max Di Cesere, a project engineer employed by Transfield. Dubois Rodney Dubois, an inspector with WorkCover Authority, the Informant. Dr Gomes Dr Lewis Gomes, a principal of Connell Wagner. Mion John Mion, a civil foreman with Transfield. NSR Project New Southern Railway Project. O'Brien Peter O'Brien, an employee of Connell Wagner. Patton John Patton, a foreman with Transfield. Dr Redman Dr Peter Redman, the principal employee of Coffey in respect of the NSR Project. Shepherd Malcolm ('Mal') Shepherd, employed by Transfield as Project Manager, Cahill Park. TBJV Transfield Bouygues Joint Venture. Transfield Transfield Pty Ltd ACN 000 854 688. WorkCover WorkCover Authority of NSW. DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.