Thiess Pty Ltd and John Holland Pty Ltd v Parsons Brinckerhoff Australia Pty Ltd

Thiess Pty Ltd and John Holland Pty Ltd v Parsons Brinckerhoff Australia Pty Ltd - [2016] NSWSC 173 - NSWSC 2016 case summary — Zoe

[4]

The real issues in dispute

The parties prepared a consolidated statement of the real issues in dispute. Some of those issues do not require decision (for example, because they were withdrawn, or because they were seen to add nothing, one way or the other, to the case of the party propounding them). When I deal with the agreed issues, I shall indicate those falling in that category.
To give content to the issues, it is necessary to know a little bit about the location of the collapse. The collapse occurred in the downdrive of MCAA, near its intersection with MC5B. MC5B was a ventilation tunnel also known as the Marden Street ventilation tunnel. MCAA was an exit ramp from one of the motorway tunnels to the Pacific Highway. The ground at that location was of low or very low strength, and intersected by a dyke (see at [52] below).
At the point where the two tunnels intersected, MC5B, which had been driven generally from southeast to northwest, swung around so that it headed almost due north. MCAA ran (or was intended to run) generally east to west.
Locations within the tunnels were identified by reference to "chainages". Chainages were measured from a set point, in increments of one metre. Thus, a location in a particular tunnel could be identified simply as (for example) "CH 100". In some documents, the letters "STN" are used instead of "CH".
One might think that the "intersection" of the two tunnels was the space enclosed by the lateral boundaries of each tunnel at the location where they met and crossed. However, the design called for the eastern and western walls of MC5B to be chamfered as that tunnel approached the point of intersection with MCAA. The parties' conventional usage treated the area to the north of the chamfers as part of the "intersection". The chamfers were at approximately 45 degrees to the intersecting walls of each of the tunnels. I set out (figure 1) a sketch showing the relationship of the two tunnels (the North bearing is an approximation, but sufficiently accurate for the purposes of these reasons): Figure 1
I set out (figure 2) an enlarged version of part of figure 1. It shows in cross-hatching the area of the intersection, strictly so-called. The greater area of the intersection, as the parties and the witnesses conventionally treated it, runs from CH 470 to CH 486.4 (approximately) in MCAA, and northwards from CH 186 to CH 198 in MC5B:

[5]

Figure 2:

With that background, I set out the parties' consolidated statement of the real issues in dispute: 1. What documents and / or oral directions comprised the relevant designs for: (a) the Intersection and its surrounds, particularly the ground support to be installed at the Intersection and its surrounds? (b) the geotechnical monitoring system to be installed at and around the Intersection and its surrounds? 2. Who prepared and provided the relevant designs referred to in Question 1(a) and (b) and when? 3. Were the relevant designs referred to in Question 1(a) and (b): (a) used or followed by TJH without any material departure? (b) provided to, and verified by, URS? Departures from design 4. In excavating and installing support to the Intersection and its surrounds, did TJH: (a) fail to comply with any of the site instructions issued by TJH? (b) fail to comply with the "cut one bolt one" policy? (c) fail to comply with any of the recommendations in any of the GSDs? (d) fail to excavate other than by way of a full face excavation? (e) fail to apply, and extend to the floor, 200mm of shotcrete where required by the relevant design? (f) fail to complete the excavation of and installation of support to MC5B before commencing the excavation of the down drive and/or up drive of MCAA? (g) fail to properly install and tension rockbolts? (h) fail to install the number and pattern of rock bolts required by the relevant design? 5. If so, were any of these failures a departure from: (a) the relevant designs? (b) the relevant designs as verified? (c) the site instructions issued by TJH? (e) any recommendations contained in the GSDs? 6. If so, were any of these failures: (a) such that they should have been apparent to TJH? (b) such that they should have been apparent to any or all of PB and / or PSM? (c) brought to the attention of TJH by any or all of PB and / or PSM? (d) if so, what did TJH do, if anything, to remedy the said departure? Contractual questions 7. Did the PB Consultancy Agreement oblige PB to undertake such inspections, supervision or reviews of the works during the excavation of and installation of support to the Intersection and its surrounds as should have made it aware of any design departures by TJH that had occurred? 8. Under the PB Consultancy Agreement, did PB have any power or authority to direct or instruct TJH, including to compel cessation of work? 9. In performing the PSM Consultancy Agreement, what was PSM obliged to do during the excavation of and installation of support to MC5B and the Intersection and its surrounds? 10. In performing the Sub Deed what was URS obliged to do in respect of verifying the relevant designs? Breach 11. Having regard to the known geology at the Intersection and its surrounds both prior to, and during, the excavation of and installation of support to the Intersection and its surrounds, and the surface monitoring equipment at the Intersection and its surrounds: (a) did PB discharge its obligations to TJH in respect of the relevant designs and PB's construction phase services including in respect of monitoring and any departures from design by TJH? (b) did PSM discharge its obligations to TJH in respect of the services to be provided by PSM during the excavation of and installation of support to the Intersection and its surrounds including in respect of monitoring and any design departures by TJH? (c) did URS, in the period to 8 October 2005 discharge its obligations in verifying (if it did verify) drawings 0024, 0059-0, 0059-1, 0061, 0072-2, 0072-3, the Geotechnical Specification and (sketch) 43-5? (d) did PB and/or PSM discharge their obligations to TJH in preparing and issuing the GSDs produced during the period between 10 October 2005 and 1 November 2005? 12. Did PB and/or PSM engage in misleading or deceptive conduct in contravention of the Trade Practices Act 1974 (Ch) in respect of the matters referred to in 11(a) and/or 11(b) and/or 11(d)? 13. Are any of the Defendants able to rely on section 5O of the Civil Liability Act 2002 (NSW) to avoid liability to TJH in negligence? Causation 14. If the answer to any or all of questions 11(a), 11(b), 11(c) and/or 11(d) is 'no', had PB, PSM and/or URS discharged its/their obligations, would the collapse have been averted? 15. If the relevant designs were provided to URS and: (a) URS had required a review of those designs as at, or no later than, 8 October 2005, and (b) that PB had conducted a review: precisely how and in what respects (analysing the matter as at 8 October 2005), and if so, why (and by reference to what) would the designs have likely changed? 16. In the event any of TJH's alleged departures from the relevant designs (referred to in Issues 4 and 5) occurred: (a) did they cause or contribute to the collapse? (b) if so, were they the sole cause of the collapse? 17. As to monitoring: (a) prior to the excavation of and installation of support to the Intersection and its surrounds, what were the instrumentation monitoring requirements as regards the Intersection and its surrounds called-up by the relevant design? (b) during the course of the excavation and support installation works at the Intersection and its surrounds what, if any, further instrumentation monitoring relevant to the Intersection and its surrounds was called-up, and by what party? (c) in each case, were those instrumentation monitoring requirements adequately performed? (d) if not, did any such failure cause or contribute to the collapse? Damages 18. In the event that TJH establishes that one or more of the Defendants is liable: (a) is TJH entitled to replacement cost of the buried plant and equipment or some other (lesser) amount, being the buyback amount or its depreciated value (Items 1 to 10 of the Scott Schedule)? (b) has TJH discharged its evidentiary onus in respect of those Makegood Losses that are in dispute (Items to 12 to 54 of the Scott Schedule)? (c) is TJH entitled to Overhead and Profit Margin as claimed in Items 56 and 57 of the Scott Schedule? (d) does the issue of contributory negligence arise at all and, if so, should any reduction be made for contributory negligence? (e) does clause 20.1 of the PB Consultancy Agreement: (i) limit the damages which TJH would otherwise be entitled to under section 82 of the Trade Practices Act 1974 (Cth)? (ii) provide a sufficient basis for the Court to decline to exercise its discretion to award pre-judgment interest to TJH under section 100(1) of the Civil Liability Act 2005 (NSW)? (f) does clause 6.1 of the URS Sub-Deed (limiting URS' liability from all claims howsoever arising to $25m in aggregate) provide a sufficient basis for the Court to decline to exercise its discretion to award pre-judgment interest to TJH under section 100(1) of the Civil Liability Act 2005 (NSW)? (g) whether TJH's claim is apportionable as against any or all of the Defendants under Part 4 of the Civil Liability Act 2002 (NSW) and, if so, what are the relative proportions of the loss and damage to be borne by each Defendant? (h) whether TJH is entitled to indemnity from URS under the Sub Deed having regard to its proper construction and the circumstances? Waiver of subrogation 19. Is TJH precluded from recovering all or part of its loss from URS and/or PSM on the basis of any waiver of rights of subrogation by its insurers under the relevant policies and, if so, what part? Cross-claims 20. In the event that TJH succeeds against any Defendant whose claim is not apportionable under Part 4 of the Civil Liability Act 2002 (NSW), is that Defendant entitled to succeed on its cross-claim against any other Defendant under section 5(1)(c) of the Law Reform (Miscellaneous Provisions) Act 1946 (NSW) and, if so, to what extent?
I note that, in relation to issues 11(c) and 15, there was a dispute between TJH and URS as to whether the temporal limitation "in the period to 8 October 2005" was appropriate. URS said that this limitation arose on a proper reading of TJH's pleaded case. TJH said that it did not. This somewhat arcane dispute does not require resolution as between TJH and PSM.

[6]

Overview of the contracts

The project was governed (as one might expect for a project of this nature) by a complex suite of contracts. The contracts may be described hierarchically in the following manner.
First, at the head of the hierarchy, there was an agreement made on 14 December 2003 known as the Lane Cove Tunnel Project Deed (the Project Deed). The parties to that deed were Lane Cove Tunnel Company Pty Ltd (LCTC), Lane Cove Tunnel Nominee Company Pty Ltd (the Trustee) and the Roads and Traffic Authority of New South Wales, as it was then known (RTA). Clause 2.1(a) imposed on the Trustee a number of fundamental obligations, including to "finance", "plan" "design" and "construct" the "Project Works and the Temporary Works…". The same clause imposed on LCTC fundamental obligations including to "operate, maintain and repair the Motorway…".
The next document down in the hierarchy is a Design and Construction Deed made in 2003 between the Trustee, LCTC, Thiess Pty Ltd and John Holland Pty Ltd (the D&C Deed). It is sufficient for present purposes to set out the recitals to that deed: A. RTA, the Trustee and the Company have entered into the Project Deed and have agreed that RTA will: (i) grant to the Trustee the right to, and impose on the Trustee the obligation to, finance, plan, design, construct and commission the Project Works and the Temporary Works; and (ii) then grant to the Company the right to, and impose on the Company the obligation to, operate, maintain and repair the Motorway (except for Asset Renewal) and to maintain and repair the Third Party Works, and levy and collect tolls, and grant to the Trustee the right to, and impose on the Trustee the obligation to undertake the Asset Renewal. B. The Trustee wishes to engage the Contractor to plan, design, construct and commission the Project Works and to carry out the Contractor's Works so as to ensure that the Trustee complies with its obligations in relation to the planning, design, construction and commissioning of the Project Works and in relation to the Contractor's Works under the Project Deed. C. The Contractor has agreed to work exclusively with the Trustee and the Company on this Project.
As is probably self-evident, the word "Contractor" is a collective term comprising both Thiess and John Holland.
The obligations of PB, PSM and URS were the subject of separate agreements, further again down the hierarchy.
TJH entered into a Consultancy Agreement with PB on 19 January 2004. Under that agreement, PB was appointed "to perform the Services". Those "Services" were said to be "all services [PB] is required to provide to [TJH] in accordance with this contract and as are further described in Schedule 3".
The services described in Schedule 3 included both design phase and construction phase services. The design services were to be prepared for "Design Lots", which were to be agreed between TJH and PB.
TJH entered into a Consultancy Agreement with PSM on 7 October 2004. That agreement was described as one "for provision of Senior Rock Mechanics Engineer and Associated Services". Under it, PSM undertook to supply or perform "the services described in Part A of the Schedule".
LCTC, the trustee, TJH and others entered into a "Subdeed of Appointment of Independent Verifier" with URS on 4 December 2003. By that deed, URS was appointed to that role. Again, it is sufficient for present purposes to set out the recitals: A. On or about the date of this Deed, the Company and the Trustee entered into the Project Deed with RTA in respect of the Project. B. On or about the date of this Deed, the Company, the Trustee, RTA, the Security Trustee and the Independent Verifier entered into the Deed of Appointment of Independent Verifier. C. Financing for the Project is being provided by the Financiers pursuant to the Debt Financing Documents. D. The Contractor is undertaking the design and construction of the Project under the D&C Deed. E. The Operator is undertaking the operation and maintenance of the Project under the O&M Agreement. F. The Independent Verifier is experienced generally in construction and project management and, in particular, in the construction and project management of major road and tunnel projects and offers its expertise in those fields. G. The Project Documents and the Debt Financing Documents contemplate that the Independent Verifier will discharge those functions set out in Schedule 2, 3, 4 and 5. H. The Independent Verifier will perform its obligations on the terms and conditions of this Deed.
The complexity (no doubt, necessary) of the contractual documentation is increased by the way in which obligations appearing in a contract at one level of the hierarchy are repeated in, so as to become terms of, a contract or contracts at a lower level. Another factor increasing the complexity (and decreasing the immediate comprehensibility) of the contract documents is their heavy reliance on a system of extensive and cross-referenced definitions. A consequence of the former characteristic is that, to understand the content of an obligation called up in one contract, it is necessary to refer to another. A consequence of the second characteristic is that to make any sense at all of the obligation thus discovered, it is often necessary to follow through a chain of cross-referenced definitions.
URS prepared a Project Verification and Monitoring Plan. The evident purpose of that plan was to describe in detail what it was that URS was required to do, in its capacity as Independent Verifier, and how it was to perform those obligations. That plan is one of those documents for which, to understand the obligations undertaken by URS so as to make sense of their intended mode of performance, it is necessary to go back to the Project Deed.
The D&C Deed required (among many other things) that TJH should comply with the obligations of the Trustee set out in the Project Deed. TJH were thereby required, in compliance with cl 5.1 of the Project Deed, to implement a quality control system for all aspects of the projects works. That was to be done in accordance with a specified section of the scope of works, exhibit A to the Project Deed.
PSM prepared for TJH, and TJH thereafter adopted and issued, a "Work Method Statement" (WMS) relating to "Geotechnical Mapping and Ground Support Determination" (the GSD WMS). The purpose of that document, as it appears from its preface, was to "[discuss] the relationship between tunnel support design and construction methods and geological classification". It was stated to be "applicable to temporary and permanent underground works of the [project] site".
PB prepared a "Final Design Report" for each of the various design lots. One related to "Design Lot: 2 Rock Reinforcement and Shotcrete Lining". That document described Dr Nagendran Loganathan of PB as the "Senior Tunnel Engineer" and Dr Doug Maconochie of PB as the "Tunnel Design Manager".
Another Final Design Report prepared by PB related to "Design Lot: 3.3 Temporary Works: Marden Street". That document described Mr Damien Leis of PB as the "Lead Tunnel Engineer", and again referred to Dr Maconochie as the Tunnel Design Manager.
It will be apparent from what I have said that in order properly to understand the obligations cast on PSM under its consultancy agreement, it is necessary to go to other documents. Those other documents to which reference must be made include not only (from time to time) contractual documents standing above the consultancy agreement in the hierarchy, but also detailed and prescriptive (and occasional proscriptive) documents such as the WMS to which I have referred.
Mr Williams of Senior Counsel, who appeared with Ms Chan of Counsel for PSM, went further. He submitted that it was necessary to bear in mind that the PSM consultancy agreement formed part of a complex interacting suite of contracts, each of which in various ways impacted on the others. That interrelationship, he submitted, arose not only from the terms of the documents themselves in some cases, but also, effectively by inference, because of the overarching requirements for the production and observance of a quality assurance and management system, and the requirement for reports and verifications at various stages, and the end, of construction. Mr Donaldson of Senior Counsel, who appeared with Mr Marskell of Counsel for TJH, did not accept Mr Williams' proposition, at least in the way it was put. I shall return to this point.
The various documents that I have briefly described (and the myriad other documents that, thankfully, were not in evidence) are of a kind commonly found in large infrastructure projects. The point was made by Dr Burman, an expert called by PSM, in the following terms (report dated 25 February 2015, page i): As is common with large civil infrastructure projects the LCT project was established under a series of contracts, deeds and undertakings that are intended to define the roles and responsibilities of the various parties at all levels together with international standards for quality assurance. At the operational level there were management plans and sub plans, communication protocols and work procedures that further defined roles and responsibilities cascading to the individual task levels. Quality systems evolved from the defence and manufacturing industries, have been part of civil engineering construction for some considerable time. Under the terms of its contract with TJH, PB was required to operate a documented quality assurance system in accordance with ISO 9001. PB developed, with TJH approval, its Design Management Sub Plan (DMSP) for the overall control of project management, quality management for its design of the LCT. The DMSP was necessarily detailed and proscriptive; it formalised tasks, roles, responsibilities, documentation, communications and processes for both design and construction stages of the project.
Whilst Dr Burman's observations on this topic might not represent the application of specialised knowledge to assumed facts so as to lead to the production of an opinion relevant to the existence of a fact in issue, they do nonetheless provide a useful summary of the way in which such projects commonly are documented. In repeating them, I do no more than acknowledge their aptitude for this task.

[7]

Relevant personnel engaged on the project

In what follows, I set out a brief description of the various personnel engaged on, or in respect of, the Marden Street Temporary Works, with an indication of their functions.
Work in the tunnels was undertaken by crews who had responsibility for individual faces, or "headings". Each crew had its own leading hand. The leading hand of the crew working in the MCAA downdrive at the time of the collapse was Mr John Labruyere. He gave evidence.
The crews were organised in two shifts: the day shift and the night shift. The day shift commenced work at about 11am and worked through to about 9pm. The night shift commenced work at about 9pm and worked through to about 7am. The period between 7am and 11am was used for repair and maintenance of the machinery, among other things.
There were no day shifts on Saturdays and Sundays. However, the Friday night shift of necessity finished on Saturday morning. The next night shift would be that starting at 9pm on the following Sunday evening.
Each shift had its own shift boss. Each shift boss was responsible for a number of crews on a number of headings. The shift boss would move from crew to crew on the different headings in the course of each shift. No shift boss gave evidence.
Standing above the shift bosses was the Site Superintendent. For present purposes, Mr John Wilson (who gave evidence) was the Site Superintendent.
All of the personnel that I have described so far were TJH employees. However, TJH had other personnel working in the tunnels. They included a Site Engineer (Mr Tim Bondin) and a Senior Project Engineer (Mr Sam Jones). Mr Bondin gave evidence. Mr Jones did not. There were of course many other engineers employed by TJH in connection with the project.
PSM was required to, and did, provide the services of an engineering geologist. For the period and tunnels in question, that role was filled principally (if not entirely) by Mr Daryl Gilchrist. Mr Gilchrist's roles included inspecting each of the excavated faces in the morning, when the night shift had finished, and mapping the observed geological conditions. Mr Gilchrist was from time to time accompanied by Dr Phil Lloyd, an employee of PB. Dr Lloyd was described as a "Senior Tunnel Engineer".
Mr Gilchrist was required to prepare documents entitled "Ground Classification & Ground Support Determination" (generally known, and in these reasons referred to, as "GSDs"). The GSDs were prepared using a standard form which provided for descriptions of: 1. the particular work site, the face chainage and the time and date of inspection; 2. the work in progress; and 3. observations of the work and the geological conditions.
The GSDs also required a statement of what was, in the opinion of the person completing them, the "Ground Class" according to the classification system to which I refer at [49] and following below. They then provided for a statement of "Recommendations" as to "Support Type". That was based upon the stated ground classification and whatever drawings might be specified.
When Dr Lloyd had accompanied Mr Gilchrist on his inspection, the GSDs noted this. They noted also, where appropriate, Dr Lloyd's concurrence in the recommendation made by Mr Gilchrist.
From time to time, Mr Jones accompanied Mr Gilchrist on the latter's inspections. Where this occurred, it was noted on the GSD. Again, if relevant, GSDs would note Mr Jones' concurrence in the recommendations as to support type.
There is an issue in this case as to whether a GSD issued by Mr Gilchrist on 17 October 2005 (see at [213] to [221] below), recommending a support type and attaching a drawing amplifying and giving content to the recommendation, is a "design" for the part of the works to which it refers.
THJ kept records of various aspects of the work as it was carried out. They included records of the number of rockbolts inserted from day to day, usually with an indication of where they were inserted. They included also records of the quantities of shotcrete applied from day to day, again usually with a record of where shotcrete had been applied. Those records were summarised, along with other matters, in "Shift Summaries". The shift summaries were supposed to be signed by the shift boss and countersigned by the site Superintendent. Usually, at least one of those functionaries did so.

[9]

The description of the defects that helped to define LCT G7 included "minor weak clayey or sandy beds, seams or joints" and "minor shear zones, faults, dykes".
A dyke in this context is a geological intrusion of initially molten material into a parent rock mass. It is generally vertical or subvertical, and generally linear. In this case, the dyke comprised dolerite material. Where the dyke first appeared in MC5B, it consisted principally of clayey dolerite material. Where it reappeared later in the intersection, it consisted principally of weathered rock material with clayey infill.
LCT G8 ground was described by listed features of the ground mass that were very similar to those describing G7 ground. The difference between G7 and G8 ground appears to relate to another characteristic, namely the spacing of discontinuities within the rock mass. Fortunately, nothing turns on this.
The characteristics of LCT G9 ground included "weak clayey or sandy beds, seams or joints, fault zones with crushed, weathered or broken rock" and "vertical or sub-vertical features such as weathered dykes and associated clay infill".
It will be seen that each classification contemplated that there might be dykes running through the rock mass. However, for G7 and G8 ground, the dykes were presumably to be such as could be characterised as "minor".
It is convenient to note at this point that three of the experts, Mr Peck (called for TJH), Dr Burman and Mr Kotze (both called for PSM), considered that the classification system lacked precision, so that "[t]he conditions in the intersection could have been assessed as any of G7, G8 or G9" (the quotation comes from the joint report prepared following their conclave).
Drawing TW03-0041-5 contains further tables. Tables 4 to 7 deal with "support class notes" for various aspects of the Marden Street Works. Table 6 deals specifically with MC5B. It prescribes different "support types" for the various ground classifications. For LCT G4, G5, and G7, the prescribed support type is MAR VII (and construction details are indicated). However, and in my view significantly, for LCT G8 and G9 ground, table 6 specifies "modified MAR VIII". By reference to drawing TW03-0043-5, MAR VIII construction involves, among other things, the installation of "steel sets": that is to say, passive steel bracing. That requirement is confirmed in table 6 of drawing TW03-0041-5, which indicated that, for G8 and G9 ground conditions, the steel sets were to be spaced at 1200mm intervals.
On or shortly before 19 August 2004, there was a rock fall in an unrelated project, the Cross-City Tunnel, then being constructed. A worker in that tunnel was killed. As a result, Mr Steve Wille of TJH issued an email dealing with "the use of spot bolts as the tunnel support regime" (this was a reference to an earlier memorandum issued by Mr David Bamforth of TJH). Mr Wille said that the requirements in Mr Bamforth's memo: … are additional to those provided by any of the tunnel geologists or details contained in the design documentation, i.e. under no circumstances can these requirements be extended by advice from geologists or design documentation, they can only reduced. 1. The use of spot bolts can only compliment [sic] pattern bolting, i.e., to pin a particular feature identified. 2. Pattern bolts are to be installed such that the maximum clear distance between the previous round on fully installed pattern bolts and the face can not exceed three metres. Subject to the Ground Classification this distance may be reduced but never extended. 3. Under no circumstances can more than one round of bolts be installed as part of a single excavation sequence. I.e. the excavation sequence is cut one, bolt one irrespective of span and ground conditions.
The third of those requirements was known, and referred to in the proceedings, as the "cut one bolt one policy". It is in my view clear that this policy was a requirement of construction practice, and not a design requirement.
The suite of designs prepared by PB included, on a drawing numbered TW03-0043-5, some four "support types". They were entitled, respectively, MAR V, MAR VI, MAR VII and MAR VIII. MAR VII is of particular relevance. It describes a support system for a tunnel that has an arched crown. The support system includes some nine rockbolts across the crown of the tunnel, and contemplates the use of additional rockbolts "as required". The notes to that design read as follows: FULL FACE CONSTRUCTION SEQUENCE (MAR VII) 1. EXCAVATE 2000 ADVANCE. AT NO TIME SHALL THE UNSUPPORTED FACE ADVANCE MORE THAN 2.0m 2. INSTALL 1 ROW OF ROCK BOLTS AND PRETENSION TO 50kN AND FULLY GROUT INSTALL SPIDER PLATES 3. APPLY 125 THICK INITIAL FRS 4. REPEAT STEPS 1 TO 3 AS PER GEOTECHNICAL ENGINEERS INSTRUCTION. OTHERWISE CONTINUE STEPS 5, 6 AND 7 IMMEDIATELY. 5. APPLY 75 THICK FINAL FRS 6. AT NO TIME SHALL THE FACE ADVANCE MORE THAN 2000 FROM FINAL FRS LAYER 7. APPLY 35 THICK FIRE PROTECTION LAYER
Drawing TW03-0059-1 contains a note 3, which states: For tunnel support notes refer to DRG no… TW03-0041
That is a reference back to the tunnel support notes that include the ground classification system to which I have referred already. The notes to that drawing seem to me to be consistent with the observational approach to design. They include the following: 3. SUPPORT DESIGN BASED ON AVAILABLE GEOLOGICAL INFORMATION AND IS SUBJECT TO CONFIRMATION BASED UPON ACTUAL GROUND CONDITIONS ENCOUNTERED … 5. SELECTION OF SUPPORT TYPE AT ANY GIVEN CHAINAGE SHALL BE BASED UPON AN ASSESSMENT OF ACTUAL GROUND CONDITIONS ENCOUNTED [SIC] BY A SUITABLY QUALIFIED GEOTECHNICAL ENGINEER 6. ACTUAL GROUND CONDITIONS ENCOUNTERED ARE TO BE ASSESSED BY A SUITABLE QUALIFIED ENGINEER GEOLOGIST AS PER TABLES 1 AND 2 … 8. SPOT BOLTING LENGTH, DIRECTION AND LOCATION SHALL BE AS DIRECTED BY A SUITABLY QUALIFIED GEOTECHNICAL ENGINEER
The reference in note 6 to "Tables 1 and 2" is a reference back to the ground classification tables (in drawing TW03-0041-5), covering among other things, the classification of shale ranging from LCT G6 to LCT G9.
Drawing TW03-0059-1 also called for an arch of shotcrete 200mm thick, applied in two layers. The construction sequence was described as follows: CONSTRUCTION SEQUENCE 1. EXCAVATE 1500 ADVANCE. AT NO TIME SHALL THE UNSUPPORTED FACE ADVANCE MORE THAN 2.5M. 2. INSTALL 1 ROW OF ROCKBOLTS WITH SPIDER PLATES AND PRETENSION TO 50KN AND FULLY GROUT. 3. APPLY 100mm INITIAL LINING FRS TO CROWN AND WALLS. 4. INSTALL 5M BOLTS (THROUGHOUT INTERSECTION) AND PRETENSION TO 50KN AND FULLY GROUT. 5. APPLY 100mm FINAL LINING FRS TO CROWN AND WALLS.
Another drawing prepared by PB, TW03-0057-0, described the excavation sequence for MC5B and MCAA. It suggested that MC5B should be completed to its northern extremity, and the tunnel roof bolted, before the breakouts into the MCAA up and downdrives were commenced.
In summary, PB prepared a portfolio of support designs for the various categories of rock that were thought likely to be encountered. PB also recommended the support types to be used in various parts of the project works and for various ground conditions.
Part of the role of the engineering geologist was to examine the ground conditions actually encountered, to classify them in accordance with the ratings that I have set out, and to record the classification in a GSD. Based on that classification, the engineering geologist was required to recommend a support type, and to record that in the GSD. If the ground conditions proved to be substantially worse than predicted, one might expect that the recommendation would specify a more capable type of support.
As a matter of practicality, it was not desirable to have rapid changes of support type. Thus, once a support type had been recommended that was thought to be adequate to the conditions, good practice would require that it continue to be used until such time as the engineering geologist was satisfied that any strengthening in ground conditions was likely to persist for some time. It would be inefficient to chop and change between support types.
Thus (among other things), it might be desirable to continue using a support type appropriate to a classification of G8 notwithstanding that the ground conditions improved to G6. If it were thought that the ground conditions would continue as G6 for some time, it might be appropriate to change the support type when the change was encountered. (In this respect, it should be noted that information obtained from within the tunnels were supplemented, as I understand it, by information obtained from an above-ground drilling program). Of course, if ground conditions became worse, there should be an immediate change to more capable support.
There were other procedures in place for requesting clarification of or changes to designs, where conditions in the ground changed. There were also other procedures in place for directing the crews as to how it was that work should proceed.
As is common in construction projects, the engineering geologists or (more usually) the TJH engineers could issue a "Request For Information" (RFI). Typically, an RFI was issued by TJH and directed to PB. It would refer to conditions encountered in the ground, or to a stage of the works about to be reached, and ask PB to advise accordingly (for example, as to some construction detail or sequence, or as to the type of support to be adopted).
Site instructions (SIs) were, as their name suggests, instructions issued to personnel on site. Typically, they were issued by one of the TJH project or tunnel engineers, and directed to the Superintendent and the shift bosses. One of the questions in this case is whether SIs issued from time to time, describing the support type to be used and (in some cases) the construction sequence to be adopted, formed part of the "designs" for the MC5B tunnel, the intersection and the MCAA tunnel. That debate focuses specifically on SIs 150, 166, 169 and 174. SIs 150, 169 and 174 were issued by Mr Bondin. SI 166 was issued by Mr Jones.
SI 169 (see at [222] to [226] below) picked up on Mr Gilchrist's GSD of 14 October 2005, (see at [213] to [221] below). SI 169 attached a revised version of the coloured drawing that Mr Gilchrist had attached to his GSD.

[10]

Method of construction of the tunnels

To assist in an understanding of what follows, I shall give a brief and (I hope) non-contentious description of the method by which MC5B and MCAA tunnels were constructed.
The experts agreed (in their joint report) that the rock mass at and around the intersection of the two tunnels comprised very low to low strength shale bedrock known as Ashfield shale. That was intersected by an extremely to highly weathered dolerite dyke. That dyke had earlier intersected MC5B. Its "strike" was approximately southeast / northwest.
The tunnels were excavated by machines known as headers. Material removed by the headers from the face of the excavation was carried away by trucks. The speed at which the headers could advance would depend on a number of factors, including the resistance of the excavated rock face.
The experts agreed (in the concurrent evidence session) that the intended primary function of the rockbolts was, in effect, to stabilise the rock mass in the roof, so as to enable the rock mass to function as a self-supporting arch. Generally, once bolting was complete, fibre reinforced shotcrete (often called FRS) would be placed in layers up to a specified depth. The shotcrete functioned both to guard against rock particles falling out of the crown of the excavation and as an arch. For the shotcrete to function as an arch, it was of course necessary that it should continue across the crown and down to the floor level of the excavation on each side.
The rockbolts were to be installed in "arrays", at intervals dictated by analysis of the prevailing geology. The designs specified the number and length of rockbolts to be installed in particular conditions, and also catered for additional spot bolts to be installed where necessary.
The cut one bolt one policy required that the excavation of a face proceed only far enough to enable an array of rockbolts to be installed across the crown of the excavated tunnel, at the correct interval from the previous array. Once this had been done, the header could continue with the next cut. That sequence was to be repeated throughout the tunnels.
In tunnels of relatively small width such as MC5B, the headers that were used could excavate the entire face in one cut. Thus, once the header had advanced the requisite distance from the previous array of rockbolts, it could be pulled back to enable the next array of bolts to be installed.
In wider tunnels such as MCAA, however, the headers were not large enough to excavate the entire width of the tunnel in one cut. It was thus necessary for the header to make two passes at the rock face. It would excavate a certain distance on (say) the right hand side of the tunnel. Generally, once that was done, the header was pulled back so that it could excavate the same distance on the left hand side of the tunnel. Again, once that excavation had been completed, the sequence of bolting could take place.
In the MCAA updrive (that is to say, the drive to the right, or east, of the intersection), excavation appears to have proceeded by way of a "full face" advance: that is to say, an advance in two passes, as just described. However, in the MCAA downdrive (to the left, or west, of the intersection), the excavation proceeded by way of a "partial face" advance. Specifically, the advance proceeded some metres further down the right hand, or northern, side of the tunnel than it did relative to the left hand, or southern, side.
The defendants all asserted that the construction method (including, in particular, the cut one bolt one policy) implicitly required construction by way of a full face advance. TJH's position on this was not always clear, but in final submissions, it became apparent that TJH did contend that construction by a partial face advance was permissible.
On 26 October 2004, a dyke was exposed in a tunnel known as MC5A. That tunnel also formed part of the Marden Street Temporary Works. The ground was classified as G9: the lowest of the strength classifications devised for the project. Mr McGhie, the senior project engineer employed by TJH, sought advice from PB as to the support method that should be utilised. Dr Maconochie recommended a support system utilising rockbolts and 200mm thick shotcrete: i.e., support type MAR VII.
Mr Gilchrist predicted that the dyke would appear in MC5B on its right hand or eastern side at about CH 70, and would continue to CH 100. That prediction was from time to time revised, to update the chainage at which the dyke was first expected to appear in the tunnel.
On 15 December 2004, Mr Gilchrist noted that the dyke was about to appear in MC5B, and thus that MAR VII support should be adopted. TJH stopped work in MC5B so that the support drawings could be revised and verified.
Work resumed in MC5B on 25 August 2005. By then, Mr Leis of PB had directed the use of "revised" MAR VII support. That was superseded on 29 September 2005 (see at [198] and following below).
There were in fact two dykes that transected MC5B. One (see at [84] above) appeared at about CH 67.3 and exited the tunnel at about CH 100. It plays no further part in this story. The second dyke is more significant. It appeared in the tunnel, nearer to the left hand side, at about CH 128. It transected the tunnel, exiting on its right hand side, at about CH 155. That dyke then re-entered the tunnel, nearer the right hand side, at about CH 180. It continued in a generally north-westerly direction into the intersection. It was observed in the cut face of the MCAA downdrive at about CH 470.
Excavation in MC5B continued up to and through the intersection with MCAA, until the north wall of MCAA was reached. The crews then "broke out" into the updrive. The northern wall, or stopend, of MC5B, appears to have been reached on about 16 October 2005. Thereafter, excavation proceeded along the updrive (to the east) for some days. On 28 October, excavation broke out into the downdrive.
On 21 October 2005 there was a substantial rock fall from the crown of the excavation in the MCAA updrive. That was attributed to a "wedge failure" of the rock mass in the crown.

[12]

Competing contentions as to causation

It is convenient to give a brief outline of the competing cases advanced by the parties as to the reasons for the collapse.
TJH said that the collapse occurred principally because the support design prepared by PB and utilised in the MCAA downdrive was inadequate, having regard to the known geological conditions. They relied on computer analysis undertaken by an expert, Dr Beck. Dr Beck's analysis was said to show, among other things, that the collapse was likely to have occurred regardless of the precise details of the construction method actually employed (that is to say, regardless of the fact that the advance proceeded on a partial face basis) and regardless of what might have been (and the defendants said were) significant deficiencies in the rockbolting and shotcreting that had been carried out in the course of that excavation.
TJH contended that the other defendants also bore some responsibility for the collapse. They said, in brief, that PSM was contractually obliged to review the designs submitted by PB and to satisfy itself that the support for which they provided was adequate given the geology that was from time to time encountered. (TJH had also asserted that PSM had, and breached, contractual obligations in relation to the monitoring of ground stability. That issue was abandoned in the course of final submissions.)
TJH said that if PSM had carried out its contractual duties properly, it should have realised that PB's support design for the MCAA downdrive was inadequate, given the known geology that was encountered. Thus, TJH said, PSM bore some responsibility for the collapse that in fact occurred.
The case against URS was based on its role as Independent Verifier. TJH said that, had URS properly performed its responsibilities in that role, it too would have realised that the support designs prepared by PB were inadequate. That applied in particular, according to TJH, when URS was called upon to verify the revised support designs that were prepared in the period leading up to 25 August 2005, when excavation in MC5B was resumed, and thereafter.
As I have said, the allegations made by TJH against PB and URS remain relevant because of PSM's case on proportionate liability.
The defendants submitted that the sole, or at least primary, cause of the failure was what they said were manifest deficiencies in the construction procedures adopted by TJH. PSM, which had assumed the carriage of this issue in an evidentiary sense, said that the evidence of TJH's witnesses who were working in the tunnels at the relevant time made it clear beyond doubt that TJH had failed conspicuously to adhere to the cut one bolt one policy. That failure was manifested, PSM said, by significant advances without any bolting or shotcreting, in both the updrive and the downdrive of MCAA. In the proved circumstances, PSM said, the immediate or primary cause of the failure was those construction deficiencies.
PSM contended that it had not been shown that PB's design was inadequate. To put it another way, PSM contended that it had not been shown that even if PB's design had been adhered to faithfully, the collapse nonetheless was more likely than not to have occurred.
TJH responded by referring to Dr Beck's modelling. They said that the modelling showed, on the balance of probabilities, that the collapse was likely to have occurred where it did even if the support design had been strictly implemented (and TJH, relying on some analysis undertaken by Dr Beck of the facts of the case, submitted that it had not been shown that the design had not been implemented at least substantially).

[14]

Relevant terms of PSM's consultancy agreement

The agreement consisted of a one page document to which there were attached what were described as "conditions of agreement". Those conditions included a "Schedule" which dealt, among other things, with the services to be provided by PSM.
Clause 3 of the conditions dealt with the engagement of PSM. By cl 3.1, PSM was engaged to provide, and agreed to provide, "the Services" (which had been defined as "the services described in Part A of the Schedule").
The parties referred also to clauses 3.2, 3.9 and 3.10. I set out cls 3.1, 3.2, 3.9 and 3.10: 3 ENGAGEMENT OF CONSULTANT 3.1 TJH engages the Consultant to provide the Services, which the Consultant must carry out and complete in accordance with this Agreement. 3.2 The Consultant must take all necessary steps to clarify, define and confirm TJH's requirements for the Services and the purposes for which the Services are required. … 3.9 TJH's client has appointed a representative known as the "Independent Verifier" to carry out certain duties. At TJH's direction, the Consultant will allow the Independent Verifier to inspect the Consultant's work at any stage of performance of the Services. The Consultant will not accept any direction or instruction from the Independent Verifier. 3.10 The Consultant, as may be required by TJH, shall co-operate with TJH's other Consultants, and do everything necessary to integrate its services with the services of TJH's other consultants.
By clause 5.1, TJH was empowered to direct changes to the Services: 5.1 TJH can instruct the Consultant to make changes to or vary the Services. No change or variation to the Services will invalidate this Agreement. The Consultant must comply with TJH's instructions.
Clause 6 dealt with PSM's obligations of "due care": 6 DUE CARE 6.1 The Consultant acknowledges that TJH has entered into this Agreement in full reliance upon the particular skill, experience and ability of the Consultant to provide the Services. The Consultant must ensure that the Services when provided are fit for the purposes stated in the Schedule (but if not so stated, the purposes to be reasonably inferred from this Agreement.). 6.2 The Consultant must exercise all the skill, care and diligence of a professional Consultant experienced in providing the Services and must carry out all responsibilities in a thorough, skilful and professional manner. 6.3 The Consultant shall ensure that the Services, including all design, drawings and other documentation produced by the Consultant, shall comply with the requirements of the Agreement, all relevant Authorities, legislation, regulations, ordinances, codes, Australian Standards, TJH's standards, as set out in this Agreement and or the D&C Deed documents and any other standards established by the Client or TJH for the works, where such standards are not in conflict with Australian Standards and Authority requirements. The Consultant shall refer any conflict between a requirement of this Agreement, the relevant standards and/or any regulatory requirement to TJH for its resolution. 6.4 The Consultant is solely responsible for the performance of the Services, notwithstanding any instructions or approval given by TJH pursuant to this Agreement. Any approval by TJH shall be in principle only and not in any way relieve the Consultant of its responsibility under this Agreement. 6.5 Without limiting the generality of the foregoing, it shall be the Consultant's responsibility to provide effective, efficient and economical solutions to satisfy the specified performance objectives and design requirements. 6.6 The Consultant indemnifies TJH against all claims, losses, actions, damages, costs and expenses arising out of or in connection with the failure of the Consultant to comply with any of its obligations under this Agreement.
Part A of the Schedule set out the services to be rendered: PART A - THE "SERVICES" Period of Engagement: 02/06/04 until completion of the Project Date of Commencement of provision of Services: 2nd June 2004 General description of the Services to be provided by the Consultant:- Provision of Senior Rock Mechanics Engineer Services to overview the construction of the Project with regard to the following points:

Interpretation of data received from tunnel & surface mounted instrumentation.
Analyse tunnel mapping and compare that to conditions described in design reports to ensure that support regimes nominated are appropriate and efficient.
Liaise with the Project designers to facilitate changes to the design to tailor it to conditions experienced based on the results of instrumentation and performance of previously installed support.
Liaison with construction engineers and supervisors regarding the design requirements.
Manage TJH's monitoring activities (specified by others) connected with the Project.
Co-ordinate geologists and engineers seconded to the monitoring team by Parsons Brinckerhoff, Coffey Geosciences and the Consultant.
Hire specialists to TJH for the monitoring activity and for other specified purposes.
Ensure that adequate records of monitoring activities are kept in an agreed format for use and consultation by TJH staff and designers.
Co-ordinate the TJH survey teams inputs into the monitoring activities. TJH survey team will provide monitoring of surface settlement, tunnel roof sag and other activities necessary or required by the D&C Deed or stakeholders.
Manage and co-ordinate the installation of monitoring equipment whether installed by TJH or subcontractors.
Provide geotechnical technicians to read all instruments installed as part of the construction of the works.

It will be noted that the services to be provided by PSM to TJH were described in very general terms.
One of the documents that was used to give particularity to those services was the GSD WMS. That was a document issued (as revision 1) by TJH on 16 February 2005. However, preceding drafts of that WMS had been prepared by employees of PSM, Mr Fowler and Ms Chan.
The scope of the WMS was set out: SCOPE This procedure discusses the relationship between tunnel support design and construction methods and geological classification. This procedure is applicable to temporary and permanent underground works of the Lane Cove Tunnel Project site.
The WMS contained a number of definitions. Although not all of them are relevant, I set them out: DEFINITIONS

JSEA - Job Safety and Environmental Assessment
TJHJV - Thiess John Holland Joint Venture
Designer - Design sub consultant - Parsons Brinkerhoff [sic] (PB)
IV - Independent verifier URS
Geologist - Engineering geologists
Senior tunnel engineer - Designer's senior geotechnical engineer as specified in PB-SP-TU-DT00-0058 - Geotechnical Instrumentation Specification.
Senior Rock Mechanics Engineer - Specialist geotechnical engineer supplied to TJHJV by sub-consultant Pells Sullivan Meynink Pty Ltd (PSM) nominated as the monitoring co-ordinator.
Project Engineer - TJHJV Engineer responsible for site works.
Geotechnical team - The geologists, senior tunnel engineer and Senior Rock Mechanics Engineer are part of the geotechnical team.
Ground Support Determination (GSD) - A two part process which a) involves the classification of ground types in accordance with design categories and b) selection of corresponding support class.
Ground Support Class (GSC) - Support ascribed to various ground types in the design.

It is convenient to note at this point that, at least for the Marden Street Temporary Works, the role of Senior Tunnel Engineer (STE) was filled by Dr Lloyd of PB, and the role of Senior Rock Mechanics Engineer (SRME) was filled by Mr Phil Clark of PSM. It is convenient to note also that Mr Gilchrist was the relevant geologist for the Marden Street Temporary Works.
The WMS then turned its attention to the responsibilities of the "Geotechnical Department". It is I think safe to proceed on the basis that this expression, although undefined, is to be read as referring to the "Geotechnical team" that had been defined a few lines earlier. The WMS stated, on the topic of responsibilities: RESPONSIBILITIES Responsibilities of the Geotechnical Department are described in Table 1 and to a lesser extent Table 2. A summary of responsibilities are presented below.

Senior Tunnel Engineer Role: Direct geologists to undertake mapping and support determinations. Ensure geologist undertake work in accordance with this WMS and JSEA's. Confirm construction meets the design intent and assess design queries at site level.
Senior Rock Mechanics Engineer Role: Review mapping, ground classification and support types recommended by geologists. Engineering Geologists Role: Answerable to the Senior Tunnel Engineer and Senior Rock Mechanics Engineer, the geologists' prime responsibilities are to map rock exposures, select ground types and provide support determinations. Works are to be undertaken in accordance with this WMS and JSEA's.

After dealing with matters relevant to safety, the environment, community issues, and "risk and opportunity", the WMS turned to the question of procedure. I set out the introductory material: PROCEDURE Introduction It is generally not possible to anticipate in sufficient detail prior to construction, all the geological conditions that will occur and the locations of such conditions. Upon exposure of the actual ground conditions, and quantification of the rock conditions, the appropriate design solution is identified for implementation, and its performance is monitored to confirm its adequacy and identify potential improvements. To ensure safe and cost effective tunnel design and construction the following must be undertaken. 1. During the investigation phase, identify and characterise the ground and groundwater conditions (or a variety of such conditions), which are expected to exist. 2. Design a suite of support arrangements and excavation procedures to suit the identified variety of ground conditions. 3. Examine the ground as it is encountered during tunnelling. 4. Confirm whether or not the encountered ground conforms to the expected conditions. 5. If it does conform, confirm whether or not the GSC and excavation procedures specified should indeed be implemented in the particular circumstances encountered. 6. Assess and evaluate the performance of the installed support and excavation procedure, by means of appropriate observations and/or instrumentation. These tasks are the responsibility of the geotechnical team and of the survey team. 7. If the ground encountered is not suitably addressed by the existing design, then the designers shall be notified and prepare a new design. It is expected that the geological team will assist the designers with this preparation where required. 8. Any modification of an existing design or preparation of a new design is to be instructed by/authorised by TJH Design Director or his delegated Design Manager allowing for appropriate review times for approval by the IV, LCTC and RTA. 9. If the installed support in any part of the tunnel does not conform, inform the area's project manager and witness rework for conformity. 10. In modifying existing designs or preparing new designs, and in assessing the application of the specified support and satisfactory performance of the installed support, interaction with the designers will be required (refer to the attached flowchart, Table 1).
The WMS continued by laying down the way in which the Geotechnical Department, or team, should carry out its work. Although the parties referred to various passages, I do not propose to set them all out. I do however note that the document made it clear that "to facilitate efficient tunnelling, it is envisaged that the GSC [ground support classification] would be maintained over an appreciable length of tunnel, rather than at short intervals".
That position was then explained on the basis that "the frequent changing back and forth between support arrangements can be very disruptive to tunnelling progress and can cause confusion". Accordingly, "when more stable ground conditions are initially encountered an immediate change is not hastily implemented. A significant amount of better ground may be required before reducing the support."

[15]

The parties' submissions

Mr Donaldson submitted that the WMS ought be regarded as a "standard established by… TJH", within the meaning of cl 6.3 of the consultancy agreement. Mr Williams did not in terms contest this proposition. I think it must be correct. What was the point of preparing the WMS, if it were not intended to help define the content of the obligations undertaken by (among others) PSM, and the services required of (among others) PSM?
Mr Williams submitted that not all the obligations set out in the WMS were obligations of PSM. He noted, correctly, that the "Geotechnical team" (or "Geotechnical Department") included persons not in the employ or under the control of PSM: significantly (in Mr Williams' submission), Dr Lloyd.
Mr Donaldson appeared to accept that the only obligations imposed on PSM by the WMS were those that were, of their nature, or by their express description or allocation, assignable (or assigned) to PSM.
Again, I think, this apparently non-contentious position must be correct. The reason why it is correct is exemplified most clearly in Table 1, where the duties broadly described in the WMS are broken down in more detail, and assigned to specific personnel (by job description). Where an obligation is assigned to a PSM employee (relevantly, Mr Clark as SRME or Mr Gilchrist as engineering geologist), the obligation must be taken to be one of PSM. Where it is assigned to someone else (relevantly, Dr Lloyd, an employee of PB, as STE), then, equally it must be an obligation of PB.
Mr Williams submitted that the agreement had to be construed objectively, so as to be given a businesslike interpretation, and taking into account the relevant surrounding circumstances. Those circumstances included, in Mr Williams' submission, the overall contractual scheme (embodying the Project Deed, the D&C Deed, and the other contracts that I have described already).
Mr Donaldson did not appear to contest those propositions, although he did submit that "the extent which interlocking contracts inform each other in terms of their construction must be limited to those interlocking contracts known as objective facts to both TJH and PSM". Mr Donaldson noted, further, that the PSM consultancy agreement was executed some time after the other documents and not contemporaneously with or shortly after their execution.
To jump ahead for a moment: I do not think that this aspect of the submissions required detailed analysis. Mr Williams' general approach to construction may be accepted. It may be accepted, also, that it is likely that TJH and PSM would have been well aware of the existence of the various contracts setting out the obligations of each of the parties to them, and thereby documenting, overall, the way in which the project was to be realised and the roles of the various parties in achieving that. At the end of the day, however, it seems to me that construction of the consultancy agreement and the GSD WMS really requires attention to the terms of those documents, considered, specifically, in the context of the assignment of responsibilities set out in Table 1 to the WMS, and the work flowchart set out in Table 2.
A key, and highly contentious, aspect of Mr Williams' submissions was that following Mr Leis' email of 29 September 2005 with its attached drawing, setting out the design for the intersection (and, Mr Williams submitted, "the Dyke affected areas"), the design was effectively "frozen". In those circumstances, Mr Williams submitted, the SRME had no function to assess the GSDs and design, to see whether the design should be implemented in specific circumstances that had become apparent.
Mr Williams relied on the evidence of three of the experts, Dr Burman, Mr Peck and Mr Kotze. Dr Burman said that in his view the design was frozen, because it was a design specifically catering for G9 conditions - "the worst possible set of conditions" - so that "it would have been impossible for… [the SRME]… to have found that the conditions did not conform". Mr Peck and Mr Kotze appeared to agree with this opinion.
Mr Donaldson submitted that the observational approach was an essential feature of the design philosophy utilised. He submitted, further, that the work of Messrs Gilchrist and Clark was an integral and indeed essential part of the observational approach. In those circumstances, Mr Donaldson submitted, any conclusion that the design had been "frozen", so as effectively to suspend some of the relevant obligations, would require far more than the opinion of an expert to substantiate it. In essence, Mr Donaldson submitted, Mr Williams was arguing that there had been a variation of PSM's obligations under its consultancy agreement. Mr Donaldson submitted that one would require very clear evidence to support this proposition, and that speculative inferences from contestable facts were insufficient.

[16]

Decision

The starting point, in relation to PSM's contractual obligations, is found in cl 3 of the consultancy agreement (see at [108] above), read in conjunction with the description of the "Services" set out in Part A of the Schedule (see at [111] above. In my view, the consultancy agreement required PSM to provide the services of the SRME for the purposes set out in Part A of the schedule. The first five bullet points are of present relevance.
As I said at [129] above, it does not appear to be in dispute, between TJH and PSM, that the GSD WMS had contractual effect, in the sense that it helped to define the nature and content of the Services to be provided by PSM. Specifically, although the consultancy agreement itself said nothing about the provision of the services of an engineering geologist, the WMS did. Mr Williams accepted that, at the level of fact, the engineering geologist's services were provided by PSM. Relevantly for present purposes, they were the services of Mr Gilchrist.
The GSD WMS stated, both in summary form and in some detail, the roles of the members of the Geotechnical team (or "Department"). I have set out the summary of those roles at [117] above. In my view, insofar as that summary refers to services to be performed by the SRME and the engineering geologist, those services were to be provided by PSM. I accept, as Mr Williams submitted, that insofar as the summary refers to the STE, those services were to be provided by PB.
The ten numbered activities (for want of a better word) described under the heading "Procedure" (see at [118] above) were, in my view, activities to be undertaken or performed by the Geotechnical team. It follows that, in deciding who was to do what, it is necessary to have regard to the more detailed allocation of functions set out in Table 1 and, to a lesser extent, Table 2. However, neither the WMS in general, nor the allocations of functions in those tables, should be taken to override the provisions of the consultancy agreement.
Reading the documents together in this way, it is apparent that, of the ten numbered activities specified under the heading "Procedure" in the GSD WMS (see at [118] above), items 1, 3 to 5 and 9 were obligations of PSM. Items 1, 3 and 4 were self-evidently obligations to be performed by PSM. Item 5 could be thought to fall within the province of either PSM or PB. However, when read in conjunction with the definition of "Services" in Part A of the Schedule to the consultancy agreement (specifically, the second bullet point set out at [111] above), that activity too seems to me to be one to be performed by PSM.
The sixth activity, in my view, was one to be performed by both PSM and PB. That follows as a matter of construction, and also in my view by reference (so far as PSM is concerned) to the first three bullet points described as part of the "Services" (see at [111] above). As a matter of construction, it is the responsibility of both PSM and PB because it is explicitly said (in the WMS) to be an activity which is "the responsibility of the Geotechnical team…". That seems to me to indicate that all the team are to contribute, according to their respective areas of expertise.
The seventh activity, insofar as it may require the preparation of a new design, is specifically directed to PB. However, the Geotechnical team (including PSM) have obligations, including (if appropriate) to notify PB that the existing design does not suitably address the ground conditions encountered, and to assist in redesign if required.
Again, the ninth activity seems to me one specifically for PSM, since it is PSM's engineering geologist who is required to make recommendations as to support, and who checks on a daily basis what is happening in the tunnels. PSM's engineering geologist must have been seen by TJH and PSM as someone uniquely suitable to perform this activity. That conclusion does not seem to me to be affected, in any adverse way, because, from time to time, Mr Gilchrist was accompanied on his inspections by Dr Lloyd (or, for that matter, by employees of TJH).
As to the tenth activity, it may be said that the primary responsibility for modifying existing designs or preparing new designs fell on PB. However, the activity specifically required the Geotechnical team to "interact" with the designers in their performance of that activity. That specifically calls up, or relates back to, the third of the bullet points (in Part A of the Schedule to the consultancy agreement) describing the services to be provided by the SRME.
PSM's pleaded case (in para 15 of its List Response to the Further Amended List Statement), asserted, among other things, that there was an implied term of its consultancy agreement: … that the general description of the Services, the personnel to be provided and the responsibilities of those personnel would be more precisely delineated and defined in documents to be disseminated by TJH which would describe the role and responsibilities of the PSM Consultants and their interrelationship with the responsibilities and roles of other employees, subcontractors and consultants working on the construction of the Lane Cove Tunnel.
The implication was said to arise "in fact": that is to say, in accordance with the principles stated in BP Refinery (Westernport) Pty Ltd v Shire of Hastings (1977) 180 CLR 266 at 283.
Consistent with that articulation of the basis on which, PSM contended, the terms should be implied, the list response set out detailed particulars.
PSM's submissions did not deal with this except to say, as a matter of conclusion rather than reasoning, that the term "is implied in fact and is necessary to ensure business efficacy". The submissions did not articulate reasons why this was so.
TJH's submissions said, with equally unhelpful brevity, that "there is no warrant for the implication of [that term]". TJH noted that there was "no explanation as to how the requirements [for implication] have been met".
One of PSM's obligations under the consultancy agreement was to cooperate with other consultants and to do what was necessary to integrate its services with the services of those other consultants (cl 3.10; see at [108] above). PSM was also obliged to do what was necessary "to clarify, define and confirm TJH's requirements for the services…" (cl 3.2; see at [108] above). Further, TJH was empowered to make changes to or vary the services to be provided by PSM (cl 5.1; see at [109] above). Finally, it was PSM's obligation to ensure that its services conformed, among other things, to the standards of TJH and other standards established either by the ultimate "Client" or TJH (cl 6.3; see at [110] above).
In those circumstances, I do not think that the term sought to be implied is, as Mr Williams submitted, necessary to give business efficacy to the consultancy agreement. Further, it seems to me, the term sought to be implied could well conflict with at least some of those to which I have referred.
It may be that, if more detailed submissions had been addressed to this point, the merits of the suggested implication (to the extent that there are any) might have become more apparent. But as I have said, the submissions were entirely conclusory. In that sense, they did not assist to explain the case that PSM sought to argue on this point.
As matters stand, I am not prepared to conclude that PSM has made good its contention that the pleaded term should be implied into the consultancy agreement.
Another issue which arises, and which may conveniently be dealt with now, is PSM's contention that the design for the intersection was effectively "frozen" on and after 29 September 2005.
PSM's case, in support of its contention that the design had been frozen as at 29 September 2005, appears to rest substantially on the opinion to this effect of Dr Burman. He argued vigorously in support of that conclusion, both in his reports and in his oral evidence. I do not think that the conclusion reflects any application of his expertise (which, as is well known and I accept, is very great indeed) to the underlying facts on which, apparently, the opinion is based. Whether or not a design had been "frozen" seems to me to depend on an analysis of the relevant contractual obligations in the light of such facts that may be proved. Thus, I do not regard Dr Burman's opinion as capable of proving that the design was in fact frozen. It follows that, to the extent that Mr Peck and Mr Kotze agreed with Dr Burman's opinion on this point, I do not regard their concurrence as adding anything relevant to it.
Nonetheless, it is necessary to consider the factual bases for the opinion. I do accept that, to the extent that an understanding of those factual bases may be informed by the input of specialised knowledge based on training, study or experience, then Dr Burman, Mr Peck and Mr Kotze are well qualified indeed to provide those inputs.
I accept, further, that an expert may be able to express a view as to whether, at the level of practicality, it would be feasible for someone in Mr Clark's (or Mr Gilchrist's) position to determine whether conditions conformed to those for which the design was prepared. But that is an aspect of the performance of their duties, and not (as it seems to me) a matter relevant to the construction of the contract.
The starting point for the argument appears to be the discovery of a dyke in MC5A in October 2004. I set out the relevant events, and the responses of PB (in particular, Dr Maconochie) at [177] to [227] below.
It appears to be a premise of the argument that a proper analysis of what happened in September 2005, in relation to the design for MC5B prepared once a dyke had appeared in that tunnel and in expectation of its persisting (or another dyke's appearing) later in the tunnel and through the intersection, must be informed by an understanding of the events of October 2004. However, as Mr Donaldson submitted, there are significant factual distinctions between the conditions encountered in MC5A in October 2004 and those encountered in MC5B from September to November 2005.
It is a further and apparently essential premise of the argument that Mr Leis' design prepared on 29 September 2005 was a design for G9 conditions - or the worst possible conditions under the LCT ground classification system. That is based on his statement that "the support design caters to the projected Dyke at the junction". As I understand it, the argument builds on this to suggest that, as a matter of necessity, the design should be taken to have catered for G9 ground conditions. However, and again as Mr Donaldson submitted, it does not follow that, regardless of circumstances, all dyke-affected ground was to be classified as G9.
Mr Gilchrist was responsible for the geological mapping of the ground conditions in MC5A in late October 2004. For the four (working) days 26 to 29 October, his mapping sheets record the presence of a dyke. However, on three of those days, he classified the conditions as G7. On the other day, 27 October, he classified the ground as G9, but changed that classification from G9 to G7 the next day.
Significantly, Mr Gilchrist's mapping described the ground that the dyke intersected as being either high strength or medium to high strength Ashfield shale. By contrast, as the experts agreed, the ground in the intersection was very low to low strength shale. Thus, leaving aside in each case the influence of the dyke, the ground strength in MC5A (immediately before the point where the dyke first appeared) was substantially better than the ground strength in the intersection of MC5B and MCAA.
The ground classification table makes it clear that the presence of a dyke does not of itself dictate that the surrounding ground must be classified as G9. Each of the G7, G8 and G9 grounds accommodates the presence of a dyke (although the first two could be taken to refer to "minor" dykes). The classification of the ground depends on a number of features, such as the number of joint sets, the spacing of discontinuities, and other matters as well as the presence or absence of a dyke.
Read objectively and in context, Mr Leis' email of 29 September 2005 does not provide support for a conclusion that the design was one for G9 conditions, simply because it stated that the design "caters to the projected Dyke at the junction".
Nor does that conclusion follow from the fact that the same design had been provided for use in MC5A when a dyke appeared in it. Once it is appreciated that the ground conditions at that point in MC5A were, to use a non-technical word, better than those in fact encountered in the MC5B/MCAA intersection, no reasonable engineer would be entitled to assume that a design said to be adequate for the former must therefore, as a matter of necessary inference, be taken to be equally appropriate for the latter.
It follows, in my view, that the proposition that the design for the intersection was "frozen" after 29 September 2005 gains no support either from the events of October-November 2004 or from the statement that "it caters to the projected Dyke at the intersection".
On what basis, then, can it be suggested that the design was frozen? More specifically, on what basis can it be suggested that the provision of the design effectively suspended the obligations of PSM (as I have found them to be) under the consultancy agreement and the GSD WMS? In my view, there is no basis either on a proper reading of the contractual documents or in the events that happened to support those suggestions.
It may be accepted that Mr Leis intended to provide a support design for the intersection. It may be accepted, equally, that he thought that the design would cater for the presence of the dyke. It simply does not follow that PSM was thereby relieved of its contractual obligations (to the extent that they were relevant). In effect, this aspect of PSM's case assumes that the contract between it and TJH was in some way varied, so as to suspend the relevant obligations until after tunnelling had progressed beyond the intersection.
Certainly, that does not appear to have been the view of those on (or in) the ground. Mr Gilchrist continued to make his daily inspections. He continued to classify the ground conditions, and to provide support recommendations. Specifically, by his GSD of 17 October 2005, he suggested (through the drawing to which I refer at [215] and following below) precisely what ground support should be provided in and adjacent to the intersection.
More fundamentally, this aspect of PSM's case suggests that the observational approach to design was set aside for the time from 29 September 2005 until (and after) the intersection had been excavated. But the design philosophy was premised on observation of ground conditions and (where necessary) modifying the support design to accommodate those conditions. Let it be supposed (contrary to the fact) that the tunnellers encountered conditions that, in Mr Gilchrist's opinion, were manifestly incapable of being supported only by rockbolts and shotcrete. Could it be suggested that, nonetheless, he had no obligation to notify TJH and PB, and to suggest that some alternative design should be prepared? Or is it to be said that, notwithstanding what (on the hypothesis that I have postulated) must have been obvious to him, and indeed to any competent engineering geologist, a support design inadequate to perform its function, should be persisted with? In each case, I think, the answer is plain. And yet, if the submissions for PSM are to be adopted, in each case it must be the other answer that is to be given.

[19]

History of designs

Mr Williams contended that the story starts in October 2004, when a dyke was exposed in a tunnel known as MC5A, and the ground affected by that dyke was classified as G9. Mr Donaldson contended that the events of October to December 2004 were irrelevant. Nonetheless, since Mr Williams relied on those events, I should set them out.
When the dyke in MC5A was exposed, Mr McGhie issued RFI 136, on 27 October 2004. The RFI sought PB's advice as to whether steel sets might not be required, or whether it might be sufficient to increase the thickness of shotcrete.
Dr Maconochie of PB replied on 29 October, stating that a support system utilising rockbolts and 200mm thick shotcrete would be appropriate. He noted that the dyke was predicted to reappear in the MC5B tunnel and said: Support type MAR VII on drawing [TW03-0043-5] shall be implemented immediately when the dyke appears.
On 18 November 2004, Dr Maconochie supplemented his response to the RFI. He said that the prescribed thickness of shotcrete: … is consistent with Ground Types G8 and G9. A complete 200mm shotcrete arch support (overall 200mm) lining provides additional support of the tunnel through the Dyke and adjacent tunnel.
I interrupt the narrative at this point to note that a "Technical Memorandum" prepared by PB in March 2005 dealt with "Ground Support Principles and Implementation". That memorandum did not suggest that shotcrete served a structural purpose. For example, it said at 2.1: Shotcrete is installed to protect the rock mass from long-term weathering affects [sic], prevent small, loosened blocks falling from between the bolts and to help direct groundwater seepage from the strip drains and to the walls.
Again, at 3.2.2, the memorandum stated that: Shotcrete, and wire mesh will be applied as necessary for the safety of the works.
That philosophy was repeated at 3.2.3, dealing with "Permanent Support".
It is convenient at this point to note that the memorandum also neatly summarised the "observational approach" to design, in a way that seems to me to be relevant to issues 1 and 2. At 3.1, dealing with "Design methodology", the memorandum stated: The support design is not based on providing a single universal design for the worst case condition but rather relies on insitu [sic] geotechnical observations and mapping to vary the support locally to address variation in conditions observed in the field and a range of support designed for these variations conditions.
Returning to the narrative: on 14 December 2004, Mr Gilchrist noted that the dyke was about to appear in MC5B. He issued a GSD of that date stating that the ground class could be expected to be G8, and that support type MAR VII should be adopted.
Work in MC5B ceased temporarily on 15 December 2004. On 17 December 2004, PB reissued drawing TW03 0043, as revision 2-10. That drawing gave effect to the support philosophy that Dr Maconochie had explained on 29 October and 18 November 2004.
On 10 August 2005, Mr Jones issued RFI 1411 to PB. It noted that work was about to restart in MC5B. It stated that as a result, because of the imminent appearance of the dyke, the support type would change to MAR VII. It requested certain data from PB.
Although it is not of great relevance, on 17 August 2005 Mr Bondin issued RFI 1479 seeking PB's approval to cut the left corner where MC5B and MCAA intersected (that is to say, the south-western corner). Jumping ahead for a moment, that approval was given by Mr Bertuzzi of PSM, who had been seconded to, and who replied on behalf of, PB.
On 18 August 2005, Mr Bondin issued SI 150. That SI instructed TJH's work crews to utilise the MAR VII support system in MC5B from CH 70 to CH 105. Mr Bondin attached a hand-drawn sketch. It provided for a 9 bolt pattern at 1250mm spacings. Mr Bondin's notes stated: 1. Excavate 2.0m advance. At no time shall the unsupported face advance more than 2.0m. 2. Install 1 row of rock bolts, pretension to 50KN and fully grout installing handle bar plates. 3. Apply 125mm fully reinforced shotcrete. 4. If ground conditions continue to be good, keep repeating Steps 1, 2 and 3. If ground Conditions are bad Spray another 75mm of Shotcrete to make the total thickness 200mm. 5. Once 200mm of Shotcrete has been Sprayed restart the cycle at Step one again. 6. At not [sic] time Can the face advance more than 2m.
Mr Bondin's SI and sketch did not fully reflect what Dr Maconochie had advised in 2004. Dr Maconochie's design required 200mm of shotcrete. However, Mr Bondin (see his third step) instructed that 125mm of shotcrete be applied, and (step 4) that it was only if ground conditions were bad that an additional 75mm of shotcrete should be applied.
On 25 August 2005, Mr Bondin issued SI 153. It modified the bolting support pattern for MC5B so that there should be 10 rather than 9 bolts in each array. His instruction concluded with the words: Everything else remains the same.
On 23 August 2005, Mr Leis replied to RFI 1411. He attached a revised version of the MAR VII support drawing and instructed that it be used in MC5B where the dyke intersected it "as specified by the Geotechnical Engineer's instruction".
Mr Gilchrist followed that instruction up in a GSD dated 24 August 2005. In it, he recommended that support type MAR VII be used, with the bolt arrays at 1 metre spacings. He stated: This support spacing has been approved by Phil Lloyd following inspection.
In a GSD dated 25 August 2005, Mr Gilchrist (having classified the ground conditions as G8) recommended support type MAR VII modified to 1 metre longitudinal spacing. His recommendation also stated that spot bolts should be installed as required, and that there should be a minimum of 9 bolts in the crown of the arch, with a further 2 bolts in the right hand side wall. Mr Gilchrist said that this recommendation was based on the "RFI response" which he understood to be an "alternative to [support by steel] sets".
Mr Gilchrist had been accompanied on his inspection by Dr Lloyd. However, he did not specifically note that his recommendations were made with Dr Lloyd's approval.
On 25 August 2005, Mr Bondin issued SI 153, which had the effect of modifying the support in MC5B from a 9 bolt to a 10 bolt pattern. It may be said that this in some way modified the MAR VII support system that Mr Leis had prescribed. However, in relation to this SI (and for that matter, in relation to Mr Gilchrist's GSDs up to this time), I do not regard the specification of additional bolts, or the specification of a lesser interval between arrays of bolts, as modifying the design, at least in any relevant way. Mr Leis' response specifically required the support to be in accordance with the recommendation of the geotechnical engineer. Mr Gilchrist made his recommendation. To my mind, Mr Bondin did no more than set out his interpretation of that recommendation.
This reading of SI 153 appears to me to be consistent with subsequent GSDs issued by Mr Gilchrist on 26 and 29 August 2005 (and later). In those GSDs, Mr Gilchrist specified that there should be 8 rockbolts in the crown and 2 to 3 in the right hand side wall. That fits in easily with the pattern that was the subject of Mr Bondin's SI 153.
In any event, nothing seems to turn on this because, on 29 September 2005, Mr Leis sent the email that attached: … the revised support detail for the MC5B/MCAA intersection at Marden Street…
Mr Leis noted in his email that "the support design caters to the project Dyke at the junction". The email then continued: The MAR-XB profile will continue through to the back (North) wall of MCAA creating a vertical stopend. For safety purposes, MC5B will be driven on 4m PB300 rockbolts with spider plates to eliminate hand bolting. 100mm initial FRS is required. To meet the span requirements, the intersection will be rebolted and constructed using 5m PB300 rockbolts and a final 100mm layer of FRS. The final FRS layer shall protect all boltheads and ensure a 200mm shotcrete arch. Beyond the intersection is covered in the DT06 design package, the nominated support is subject to Geotechnical Engineers assessment. The change in support to the MAR-XB profile does not change the current MX data being implemented by Surex. Please review and comment accordingly. Cheers, Damien
The drawing attached was TW03-0059-1. It was stamped, in accordance with Mr Leis' email: CLIENT REVIEW 28 SEP 2005
The drawing called for arrays of 9 rockbolts across the crown of the arch, and a further 3 rockbolts on the right hand side of the arch. Those arrays were to comprise 4m bolts at 1m intervals, and 5m bolts at 1.5m intervals. It called further for "spot bolts to both sides as specified by geotechnical engineer", and for a 200mm thick shotcrete (in the drawing, called "FRS") lining to be applied in two layers, each of 100mm.
For convenience, I repeat the construction sequence to be adopted: CONSTRUCTION SEQUENCE 1. EXCAVATE 1500 ADVANCE. AT NO TIME SHALL THE UNSUPPORTED FACE ADVANCE MORE THAN 2.5M. 2. INSTALL 1 ROW OF ROCKBOLTS WITH SPIDER PLATES AND PRETENSION TO 50KN AND FULLY GROUT. 3. APPLY 100mm INITIAL LINING FRS TO CROWN AND WALLS. 4. INSTALL 5M BOLTS (THROUGHOUT INTERSECTION) AND PRETENSION TO 50KN AND FULLY GROUT. 5. APPLY 100mm FINAL LINING FRS TO CROWN AND WALLS.
The notes referred to a number of other drawings for various details, including the tunnel support notes in drawing TW03-0041-5. The notes also stated: 5. ASSESSMENT OF ACTUAL GROUND CONDITIONS ENCOUNTERED AND NOMINATION OF SUPPORT TYPE AT ANY GIVEN LOCATION TO BE BY A SUITABLY QUALIFIED GEOTECHNICAL ENGINEER
Another feature of the design in drawing TW03-0059-1 is that, apparently, it extends beyond the crown, and eastern and western walls, of MC5B. The drawing showed in section MC5B at the point of intersection with MCAA, and the roof and floor lines of MCAA.
Among other things, the design called for 5m rockbolts to be installed at 1500mm spacing "Bolting Throughout Intersection or as Specified By Geotechnical Engineer", apparently in addition to the bolts referred to at [201] above. The area of the intersection to which this latter direction applied appears to have extended into the area of the breakouts into the MCAA updrive and downdrive.
By scaling off the drawing, that instruction appears to extend a little under 3 metres to the west (in the downdrive) and a little over 3 metres to the east (in the updrive). That would take it from about CH 472 in the downdrive, to somewhere between CH 478 and 479 in the updrive.
On that analysis, the direction appears to relate in part to the extended concept of the intersection which, by reference to figure 2 set out at [11] above, could be taken to extend from CH 470 in the downdrive to CH 488.4 in the updrive (and, travelling south, to CH 186 in the MC5B tunnel).
A precise understanding of Mr Leis' email and the attached drawing is very difficult to achieve. On the one hand, as I have said, his directions appear to extend beyond the intersection strictly so called. On the other hand, the email calls for a 200mm shotcrete arch. That could not have been a permanent support feature, because the breakouts into the updrive and the downdrive would necessarily remove the shotcrete. (One might think that the breakout into the updrive would also remove any rockbolts installed on that side.)
At about this time, Mr Jones turned his mind to engineering problems that would arise when the excavation of MC5B entered the intersection. MCAA was steeply graded (falling from east to west). It was important that the excavation of MC5B, at the intersection, not go too low, so as to get below the design floor levels of MCAA at that point. Accordingly, Mr Jones proposed a "benching" sequence.
In this context, a "bench" denotes a horizontal layer of rock, from wall to wall of the tunnel being driven. The requirement to proceed by way of benching means that successive passes of the header would be necessary to get the floor levels in the updrive and the downdrive down to the design levels. Mr Jones indicated how this was to be done in his SI 166, dated 5 October 2005. He attached drawings to that SI which set out in diagrammatic form his requirements, and also the construction sequence to be adopted.
The first drawing that Mr Jones attached was a marked up version of TW03-0059-1. Mr Jones proposed that MC5B be driven to its northern extremity, CH 198, and that the header then be pulled back so that the excavated tunnel could be bolted. Notwithstanding PB's requirements, as set out in Mr Leis' email of 29 September 2005, Mr Jones specified 5 metre bolts on a 1.5 metre grid from CH 185 (entry into the intersection) to CH 198 (the northern wall of MC5B). Once that was done, the operations of benching, and breaking out into the updrive and downdrive, were to take place.
Mr Jones' drawings did contemplate that shotcrete would be installed, in two layers each of 100mm. However, it is clear that this was to be for temporary support only, since (in the intersection) shotcrete would be removed from the walls once the breakouts east and west commenced.
On 17 October 2005, Mr Gilchrist inspected the excavated face of MC5B at about CH 198. He noted (correctly) in his GSD that this was the "back wall", or north wall, of the tunnel. The GSD classed the ground conditions as G7/G8. It recommended ground support in accordance with drawing TW03-0059-1; that is to say, in accordance with the MAR VII (modified) support the subject of Mr Leis' email of 29 September 2005 with its attached drawing (see at [198] and following above.)
Mr Gilchrist then noted the way in which the excavation would proceed: - following the installation of 5m bolts I understand that a bench of ~ 1-1.5m is to be removed. - Road header will then turn RIGHT to cut up slope into MCAA. Initially the bolts are to be 4m long at 1m x 1m grid with 200mm of shotcrete down to the floor until ~ CH486 where the shotcrete reduces to 50 initial / 125 total. When road header continues "down" MCAA drive i.e.) left turn, the initial support remains @ 4m bolts in 1m x 1m grid with 200 shotcrete in sequence until ~ 469.5 () - 5m bolts will need to be installed between CH 469.5 → ~ CH486. - NOTE ( TO BE ASSESSED DAILY BY GEOTECH ENG/GEOLOGIST)
Mr Gilchrist attached a drawing to that GSD. It was a version of a drawing known as TW03-0058-1, marked up in colours. That drawing shows the intersection and its surroundings, from about CH 180 in MC5B through to the north wall, and from about CH 467 to about CH 493 in MCAA. Mr Gilchrist marked up the drawing to show: 1. the projected path of the dyke through the intersection; 2. a green area in and near the intersection (in the strict sense); 3. to the left of that green area, a yellow area with, to its left, a pink area; and 4. to the right of that green area, another pink area.
The green area extended from CH 186 to CH 198 (the north wall) and from about CH 471.5 on the western side to about CH 487 on the northern side.
The yellow area extended from CH 470 to the point where the green area started on the west. The pink area extended westward from the yellow area and eastward from the eastern boundary of the green area.
Mr Gilchrist inscribed a "legend" on the drawing, which gave content to the colour coding that I have described. The green area indicated that there were to be initially 4m bolts at 1m centres with 200mm of shotcrete, and then 5m bolts at 1.5m centres.
The yellow area indicated that there were to be 4m bolts at 1m or 1.25m centres ("(DEPENDANT [sic] UPON GROUND CONDITION]"). Shotcrete was to be applied, starting with an initial layer 50mm thick, up to a total thickness of 125mm.
In the pink area, Mr Gilchrist indicated that there were to be 4m bolts at 1.25m centres. Shotcrete was to be as per the yellow area.
Mr Gilchrist added a note to his legend: THIS SUPPORT PLAN DRAWN BY D GILCHRIST (GEOLOGIST) FOLLOWING REQUEST BY S JONES (SITE PROJECT ENGINEER) AS DISCUSSED + APPROVED BY P LLOYD (SNR TUNNEL ENGINEER)
Also on 17 October 2005, Mr Bondin issued SI 169. As its title stated, it was intended to describe the "Support Regime Through MCAA, MC5B Intersection". Mr Bondin described that support regime by reference to an attached diagram. The diagram was in effect a reworking of Mr Gilchrist's drawing based on TW03-0058-1. Mr Bondin made a number of changes.
First, Mr Bondin eliminated the yellow zone shown on Mr Gilchrist's drawing and extended the green zone westwards to CH 467. He also extended the green zone eastwards to CH 487.
Second, Mr Bondin varied the construction sequence to be followed in the green zone. He maintained the bolting patterns (4m at 1m and 5m at 1.5 metre) set out in Mr Gilchrist's drawing, but specified that the bolts were to be installed in the following sequence: 1. the 4m bolts; 2. followed by 100mm of shotcrete; 3. followed by the 5m bolts; and 4. followed by a further 100mm of shotcrete.
Mr Bondin made no change to the bolting or shotcrete recommendations that Mr Gilchrist had described for the pink zones in his drawing.
In short, the effect of Mr Bondin's drawing was to vary Mr Gilchrist's support recommendations by extending the zone within which the double arrays of bolts were to be installed and by varying the sequence for the application of shotcrete.
It is clear that both Mr Gilchrist's drawing and Mr Bondin's revision of it extended (to the west) beyond the location (near CH 467) of initiation of the collapse that occurred. To put it another way, each drawing described the support regime (or regimes) to be utilised at and beyond the point where, a little over a fortnight later, the collapse occurred.

[21]

Decision

In my view, it is necessary to start by considering the roles of the parties. TJH were the contractors, responsible for the design and construction of the project. They subcontracted design responsibility to PB.
PSM was the geotechnical consultant. Its obligations were those that I have summarised at [174] above. In that context, whatever may be the proper way to regard the distribution of responsibilities set out in the GSD WMS, the inescapable fact is that the "Services" that PSM undertook to supply, by cl 3.1 of its consultancy agreement with TJH, included the provision of the services of an SRME "with regard to", among other things, the "points" that I have set out at [111] above.
Further, the way in which the project worked (at least, in relation to the Marden Street Temporary Works) was that: 1. TJH performed the construction work; 2. they did so in accordance with designs provided by PB; 3. where clarification was needed, or where changes in conditions occurred, TJH would issue RFIs for PB to address; 4. once PB had dealt with an RFI, TJH were required to adopt any recommendation made as to change of design (in principle, after verification by URS); and 5. the designs from time to time issued by PB provided, either expressly or by cross-referencing back to other drawings, that PSM, as geotechnical consultant, had the responsibility for selecting the support type at any given place based upon its assessment of actual ground conditions.
Thus, PSM's obligations and hence its role, extended beyond assessing ground conditions, to selection of support type where necessary to suit the ground conditions actually experienced. That seems to me to be entirely consistent with that part of PSM's Services which required the comparison of actual conditions (ascertained through tunnel mapping) with conditions described in the design reports "to ensure that support regimes nominated are appropriate and efficient".
True it is that changes in design were to be undertaken by PB (although the Services to be provided by PSM included dealing with PB to facilitate design changes). However, in the present context, what was required after 29 September 2005 was either implementation of the design specified in Mr Leis' email or dealing with PB if, in PSM's analysis, actual ground conditions required something different.
Thus, it seems to me, Mr Gilchrist's GSDs, with their recommendations of ground support, should be regarded as implementing the design specified by Mr Leis in the email of 29 September 2005 and its attached drawing. However, in my view, nothing in that email or in what happened thereafter operated to change in any material way, let alone to diminish or circumscribe, the nature, quality and extent of the Services to be provided by PSM under its consultancy agreement with TJH.
When one turns to Mr Leis' email and drawing, two things become apparent: 1. it is dealing with MC5B at and through the point of that tunnel's intersection with MCAA; and 2. it prescribes two different bolting regimes, one for the crown and eastern wall of MC5B, and another extending beyond the western and eastern walls of that tunnel (see at [198] above, where I describe this dual regime).
The drawing does not show how the breakouts east and west are to be handled, or what is the intended support regime for MCAA either east or west of the eastern and western walls of MC5B, except insofar as the extended bolting regime may be thought to cover this point. That is clear from the fact that the drawing shows a 200mm shotcrete arch for MC5B: something that would, necessarily, disappear (in the form shown on the drawing) once the breakouts commence.
Accordingly, as it seems to me, Mr Gilchrist's GSD of 17 October 2005 must be regarded as taking Mr Leis' design of 29 September 2005 a step further than it was intended to travel: indeed, a step further in each direction, east and west. Mr Gilchrist appears to have regarded the 29 September design as intended to apply not only in the intersection strictly so called, but also in the extended intersection, including at least the eastern and western chamfers and the area of MCAA to their north (from CH 470 to approximately CH 488.5). Mr Bondin's SI 169 appears to have interpreted the 29 September design in the same way (or, at least, to have acquiesced in Mr Gilchrist's apparent interpretation of it).
It follows in my view, that the correct analysis of what occurred is that the designs for the intersection comprised: 1. for the intersection strictly so called (i.e., the area from CH 186 to CH 190 and from approximately CH 476 to CH 485), Mr Leis' email of 29 September 2005 and the attached drawing; and 2. from 17 October 2005, and for the extended area of the intersection, Mr Leis' design as interpreted (or modified) by Mr Gilchrist's GSD of 17 October 2005 and its attached drawing, varied the same day by Mr Bondin's SI 169 and its attached drawing.
There was a dispute as to whether the changes made by Mr Bondin to Mr Gilchrist's recommended support regime were significant. In one sense, those changes were not. Mr Bondin extended the area of closer bolting both to the east and to the west, compared to that indicated on Mr Gilchrist's drawing. I do not think that Mr Williams was bold enough to suggest that this change was deleterious.
However, Mr Williams did draw attention to the variation to the shotcreting regime to which I have referred at [226] above. Mr Gilchrist specified that, in his green zone, 200mm of shotcrete was to be applied after the 4m bolts had been installed at 1m centres, and that thereafter the 5m bolts were to be installed at 1.5m centres. Mr Bondin varied this by providing that in his (extended) green zone, the shotcrete was to be applied in two layers, one after the 4m bolts were installed and the second after the 5m bolts were installed.
In each case, the thickness of the finished shotcrete layer was the same: 200mm. In each case, shotcrete to that thickness would need to be applied in two steps (because, as everyone agreed, it was impossible, in a practical sense, to apply 200mm of shotcrete in one application). Thus, the only difference between the two sequences is that on Mr Bondin's sequence, the shotcrete would have a thickness of 100mm until the second layer was applied, which in turn would be after the 5m bolts had been installed.
Although Mr Williams said that this was an important change, his submissions did not explain why it was important. Nor, so far as I know, did any of the experts venture an opinion either as to the importance of the change or as to the potential risks of proceeding as Mr Bondin had indicated rather than as Mr Gilchrist had indicated. Although Mr Bondin was cross-examined on his drawing, it was not suggested to him that the change could have had any deleterious impact on the safety or stability of the excavation.
In my view, the evidence does not support a finding that to the extent that Mr Bondin's SI 169 varied the support regime described in Mr Gilchrist's GSD of 17 October 2005 with its attached drawing, that change had any material significance. Certainly, Mr Gilchrist does not appear to have thought so, if his GSDs of 27 and 28 October 2005 are any guide to his views. I deal with these at [271] to [275] below.

[25]

Outline of the disputes

PSM contended that the collapse in the MCAA downdrive on 2 November 2005 was caused by significant failures in design (on the part of PB) and construction (on the part of TJH). In broad outline, the construction deficiencies, according to PSM, comprised a failure to install the support system for which the design called: namely, arrays of rockbolts at specified intervals, and a 200mm thick shotcrete arch.
This aspect of PSM's case may be summarised as follows (taken from its written closing submissions at [3]): The support system consisted of rock bolts and shotcrete. The intended support was not installed adequately. The rock bolts were not installed in the design pattern and number. The rock bolts suffered from tensioning issues. The shotcrete was inadequate in terms of quantity and spread. The shotcrete was also not installed floor to floor.
PSM's case as to the alleged deficiencies in the installation of rockbolts and the application of shotcrete, was based on records prepared by TJH and by Mr Gilchrist. TJH's records showed, for each shift, the advance made on each heading, the quantity of rockbolts installed, in most cases the general areas where the rockbolts were installed, and the quantities of shotcrete applied. Mr Gilchrist's records confirmed the distance advanced on each heading, and from time to time made comments about bolting and the like.
As to the alleged deficiencies in the tensioning of the rockbolts, PSM relied primarily on the evidence of Mr Kotze. Mr Kotze had been called in (by WorkCover) immediately after the collapse had occurred. He inspected the area of the collapse on 3 November 2005: the day following the collapse. He examined a number of rockbolts that had been removed from the fallen material and conveyed to external premises for examination.
TJH accepted that their relevant witnesses had conceded, on the basis of the records put to them in cross-examination and assumptions based on those records, that there appeared to be serious departures from the design intent, relevant SIs, the cut one bolt one policy, and support recommendations contained in relevant GSDs: failures of the kinds described in paras (a), (b), (d) and (h) of issue 4.
However, TJH contended, the analysis of the records (in particular, the analysis of the rockbolt and shotcrete records) was simplistic as to the downdrive, and that, properly understood, those records did not reveal (in the downdrive) the deficiencies in number or quantity that the witnesses had been asked to assume (comparing the number of bolts, or the quantity of shotcrete, to the distances advanced from shift to shift).
TJH submitted that Mr Kotze's opinions as to deficiencies in the anchoring and tensioning of the rockbolts, were not persuasive, particularly in light of the evidence (as to what actually happened) of Mr Wilson, the Superintendent at the time.
The evidence, in relation to the various matters the subject of issue 4, is both substantial in extent, unwieldy, and difficult to analyse. That is because it comes from different sources, none easily reconcilable with the others. I propose to deal with the paragraphs of issue 4 by grouping them under various headings.

[27]

What the documents show

PSM relied on documents relating to the updrive as well as the downdrive. The relevance of the alleged deficiencies in the updrive is not immediately apparent, given that the collapse occurred in the downdrive. PSM sought to justify the relevance of those records by submitting that they were evidence of pervasive or widespread construction failures.
The records on which PSM relied were summarised by it in a table which, if I may say so with respect, made it extraordinarily difficult to extract the relevant data so as to understand the point being made. The difficulty was not diminished by the liberal application of splashes of different colours, which apparently were intended (but failed) to facilitate an understanding of the material. Nor was the difficulty alleviated by the broad-brush approach taken by PSM in its submissions.
Further, PSM's tabular summary commenced with the day shift on Monday 10 October 2005. However, the work did not break out into the MCAA updrive until the day shift nine days later, on Wednesday 19 October 2005. And it did not break out into the downdrive until some eight days later again, in the night shift on Thursday 27 October 2005.
TJH summarised the data (as to metres advanced, numbers of rockbolts used, and quantities of shotcrete applied in the downdrive) in a table that formed part of their written closing submissions. Since PSM, in its written submissions in reply, did not quarrel with the accuracy of that table (so far as it went), I proceed on the basis that (so far as it goes) it may be accepted. Accordingly, I set it out: Date and shift Metres advanced in MCAA down drive Bolts required Bolts installed 27 October 2005 (n)45 1.0 metre 9 27 28 October 2005 (d)46 4.4 metres 36 5 30 October 2005 (n)47 1.7 metres 5 (half face advance) 4 31 October 2005 (d)48 1.0 metres 5 (half face advance) 13 31 October 2005 (n)49 0.5 metres 0 0 1 November 2005 (d)50 ? ? 12 TOTALS 8.6 metres 55 61

[28]

The question marks for the day shift of 1 November 2005 reflect deficiencies in the records of metres advanced and, hence, the number of bolts required. It may be, in consequence, that the twelve bolts installed on that day were by way of "catch-up" on some pre-existing shortfall. PSM took the position that this was the correct analysis, and said it was "too little, too late".
Any attempt to understand the data is confused further by the evidence of construction practice. TJH had put some bonus system in place, designed to encourage the workforce to execute the tunnelling works as quickly as possible, consistent with the requirements of the designs and the overarching interests of safety. However, a particular crew could not "claim" metres advanced during its shift unless it had completed all works relevant to that advance, including rockbolting (and, it may be, shotcreting). Accordingly, if a crew had advanced a certain distance, but had not completed those works before its shift finished, the incoming crew could complete those works and claim the metres, before proceeding with the next cut.
Whilst I do not suggest that these features of the construction practices followed would have any distorting effect on results considered over a lengthy period of time, that may not be so if attention is to be focused on the six shifts that worked on the MCAA downdrive over the period from 27 October 2005 (night shift) to 1 November 2005 (day shift).
The analysis in the downdrive is further complicated because the excavation proceeded by way of a partial face advance. Thus, whilst a full array of (say) nine bolts might be required for every metre of full face advance, some lesser number of bolts would be installed for a partial face advance. Again, because it was a partial face advance, not nearly as much shotcrete would be applied as would be the case in a full face advance for the same distance. Indeed, it may be thought that there was little utility in applying any shotcrete until the face had been squared up (although I acknowledge that considerations of safety might dictate that at least some shotcrete should be applied to the crown, to prevent loose material from falling).
Be all that as it may, the results of TJH's analysis of the data for the downdrive show a maximum shortfall, of bolts installed compared to bolts required, of six. And if (which is entirely unclear) the advance on 28 October was a partial face advance, then there would be no shortfall, or at most a shortfall of one bolt.
Assessment of the sufficiency of shotcrete applied in the downdrive, by comparing the metres advanced with the quantities of shotcrete used, is further complicated by directions given by Mr Gilchrist and acted upon by Mr Bondin. The support design called for 4m rockbolts to be installed on a 1m grid pattern (i.e., at intervals of 1m both laterally - across the crown - and longitudinally - along the tunnel), and for 5m rockbolts to be installed on a 1.5m grid. That is the combined effect of Messrs Gilchrist's and Bondin's interpretations of the modified MAR VII design proposed by Mr Leis on 29 September 2005. It is evident in Mr Gilchrist's GSD of 17 October 2005 and Mr Bondin's SI 169 (see at [213], [222] above).
Mr Gilchrist's GSD called for 200mm of shotcrete to be applied after the 4m bolts were installed and before the 5m bolts were installed (at least, in the "green" zone of the intersection). Mr Bondin's SI modified this by providing for 100mm of shotcrete to be applied after the 4m bolts were installed and for the second 100mm of shotcrete to be applied after the 5m bolts were installed. (Mr Bondin also extended the "green" zone, but nothing of present moment turns on this.)
Mr Gilchrist's GSD of 27 October 2005 recorded among other things his observations of what was happening in the updrive. It classified the ground as G7. It made a recommendation for the support to be utilised in the updrive. However, presumably because the breakout into the downdrive was imminent, Mr Gilchrist also dealt with that. He said: MCAA backdrive (LHS) - dyke affected ground - 4m long bolts at 1m centres and handle face plates - 100mm shotcrete (min) in cycle - 5m [bolts] to be installed later - Total 200mm shotcrete thickness (min) to continue
Mr Bondin picked this up in his SI 174, dated 27 October 2005. He said, for the MCAA downdrive, that the work crews were to: Maintain 4m long bolts at 1m centres and handle face plates. Maintain 100mm minimum shotcrete in cycle.
Mr Bondin's SI did not refer to the 5m bolts, presumably because he recognised that Mr Gilchrist had said they were to be installed later.
Mr Gilchrist reported on progress in the downdrive in his GSD of 28 October 2005, by which time excavation had progressed about 1m into the downdrive. He noted that there was a cut of approximately 1m, and that two rows of bolts had been installed. He noted further that: The drive is only being cut narrow [sic] at present with LHS still to be removed. Ground currently appears to be tight with weathered dyke maintaining integrity.
Mr Gilchrist classified the ground conditions as G7/G8. His recommendation as to support in the MCAA downdrive was (in substance, although with some minor and irrelevant verbal changes) identical to the recommendation in his GSD the previous day. It did not refer to "dyke affected ground", but since this had been dealt with in his observations, nothing turns on that omission.
In short, in the period leading up to the collapse, Mr Gilchrist: 1. recognised that the advance in the downdrive was proceeding by way of partial face advance; 2. recommended installation of 4m rockbolts followed by 100mm of shotcrete "in cycle"; 3. authorised deferral of the 5m rockbolts (without specifying when they were to be installed); and 4. effectively authorised deferral of the application of the second layer, a further 100mm, of shotcrete.
I have referred already to the complexity of the information presented by PSM, in tabular form, dealing with rockbolts and shotcrete in MC5B and MCAA. I have referred also to the failure of their written submissions to address in detail the implications of that information, or to spell out in a comprehensible form the quantitative deficiencies said to be exhibited. I note as well that there was no attempt to grapple with the consequences (in terms of any asserted quantitative shotcrete deficiency) of Mr Gilchrist's recommendations contained in the two GSDs to which I have referred.
In those circumstances, I am not prepared to conclude, on the basis of the ill-presented, undigested and unanalysed material propounded by PSM, that there was any significant quantitative deficiency in the installation of rockbolts or the application of shotcrete in the downdrive. Nor do I think that this conclusion can be supported (in the way put to TJH's witnesses) by simple comparisons of metres against quantities.
Further, and dealing for the moment with a different topic, namely the complaint that the advance in the downdrive proceeded by way of partial face advance rather than full face advance, it is apparent from the GSD of 28 October 2005 that Mr Gilchrist was well aware of this. He did not suggest that it was inappropriate to proceed (so far as had been done) by way of a partial face advance, with (as he recorded) "LHS still to be removed".
Mr Gilchrist did not give evidence (although PSM had served an affidavit sworn by him). I conclude, on the basis of the GSD, that he did not regard the failure to proceed by way of full face advance in the MCAA downdrive as a construction deficiency, or as a departure from the design. In fact, there is no evidence that it did represent a departure from the design. I repeat that the machinery being used could not proceed, in a single cut, by way of full face advance in either the updrive or the downdrive.
Returning to the topic of rockbolts and shotcrete: the position (as to deficient construction practice) is much clearer in the MCAA updrive. The tunnel face there was squared up, metre by metre, so that (although each metre of advance had to be made in two cuts), it was effectively a full face advance. Thus, assessment of the metres excavated against the number of rockbolts used and the quantities of shotcrete applied is far less problematic than it is for the downdrive. Nor is there any complication arising from recommendations, as to deferring installation of the 5 metre bolts and application of the final layer of shotcrete, of that kind that Mr Gilchrist made in respect of the downdrive.
In the case of the updrive, it is much easier to conclude, and I do conclude, that there were very serious deficiencies in the number of rockbolts installed and the quantities of shotcrete applied. It is easy to conclude, and I do, that these represented serious departures from the design. It is also easy to conclude, and I do, that these deficiencies (in relation to rockbolts) represented serious departures from the cut one bolt one policy. However, in relation to that last point, the cut one bolt one policy was a construction policy adopted by TJH (for very good reasons). It was not an aspect of the design.
There is one more matter of fact to which I should refer before turning to the lay and expert evidence. That concerns deficiencies in the application of shotcrete in the MC5B tunnel.
As I have noted at [90] above, there was a substantial rockfall from the crown of the MCAA updrive on 21 October 2005. That was the result of a "wedge failure". Mr Gilchrist noted it in his GSD of 21 October 2005. The failure occurred at about CH 495. It was some metres long and some fewer metres wide, although irregularly shaped.
Perhaps not surprisingly, the rockfall caused concern. It prompted, among other actions, a check on the depth of the shotcrete in MC5B.
To that end, Mr Bondin issued SI 172 on 21 October 2005. It required the drilling of check holes from CH 160 (which is just after the dyke that later reappeared in the intersection had exited the right hand side of MC5B) to the north wall (CH 198) across the crown and in the sidewalls, according to a pattern that he illustrated diagrammatically. He said that the holes should be drilled every 5 metres (longitudinally).
Those checks were (to some extent) made. The results were referred to in SI 173, issued by Mr Jones on 22 October 2005, with an attached diagram.
Mr Jones said:

The final shotcrete thickness in MC5B is 200mm. Please refer to the attatched [sic] diagram for those areas where shotcrete is not currently not 200mm thick. The thickness needs to be increased as shown on the attatched [sic] diagram.
INSTALL DEPTH GUAGES [sic] at all locations prior to shotcreting.
WHERE SHOTCRETE THICKNESS IS ALREADY 200MM SIMPLY COVER THE BOLT TAILS (CUT OFF LONG ONES FIRST).

The diagram attached to Mr Jones' SI summarised, by showing at various locations the further applications of shotcrete required, the results of the check that Mr Bondin had directed. However, and inconsistently with Mr Bondin's instructions, the diagram starts at CH 130. Of the 66 areas indicated in the table (which extended only up to CH 185), 28 required applications of shotcrete ranging in thickness from 150mm (at two locations) to 40mm (at one location), with varying other thicknesses required at various locations.
A further 15 locations had a question mark: presumably, either there were no results, or the results did not enable any assessment to be made of the deficiencies in shotcrete thickness.
The remaining 23 locations were shown as not requiring any further application of shotcrete.
The evidence does not disclose why it was that the checks were undertaken from CH 130 to CH 185, rather than from CH 160 to CH 198. Further, to the extent that the table represents the number of checks undertaken at each chainage, it would appear that there were six checks made, rather than the eight that were referred to in Mr Bondin's SI.
Whatever may be the reasons for these discrepancies, it is clear that the checks that were undertaken indicated significant deficiencies in the application of shotcrete in MC5B. There is no reason, in the evidence, to think that the attention to detail of the work crews, or the thoroughness of their applications of shotcrete, might have improved markedly from CH 185 to CH 198 (or, for that matter, in the updrive and the downdrive thereafter).
Another matter that appears from Mr Bondin's SI is that, although shotcrete depth gauges were required to be installed as part of the work sequence (so that the shotcreters would know how much shotcrete to apply), this had not been done. Mr Bondin gave evidence on this topic. He suggested that the depth gauges were not required, because the shotcreters could use the handle plates of the bolts as a guide to the thickness of shotcrete required. Putting it neutrally, that evidence did not convince me that the work practice apparently followed (up to 21 October) provided a reliable or acceptable substitute for the required use of depth gauges.

[29]

The evidence of the lay witnesses

The evidence of those who had worked in the tunnel was uniform in recognising that, comparing the metres advanced with the quantities of bolts used and quantities of shotcrete applied in the MCAA updrive, there appeared to be serious deficiencies. The witnesses were: 1. Mr Labruyere (the leading hand for one of the crews working in MC5B and MCAA); 2. Mr Bondin; and 3. Mr Wilson (the tunnel superintendent).
There is no need to go to the minutiae of the evidence. Each of those witnesses was confronted in detail with the documents that showed the metres advanced, the numbers of rockbolts installed and the quantities of shotcrete applied. Each agreed, in relation to the MCAA updrive, that on the basis of those records there were construction deficiencies that, as Mr Williams put it, involved serious departures from: 1. the design intent; 2. relevant site instructions; 3. the cut one bolt one policy; and 4. support recommendations in relevant GSDs.
On the basis of the material put to him, each witness was right to concede those matters. I should say, in case this might be thought to imply some criticism, that none is intended. Each of them seemed to me to be concerned to tell the truth to the best of his ability, and to the extent that, ten years after the event, his memory permitted. None of them sought to dissemble or prevaricate, or to avoid the point of any particular question. I am satisfied that each of them should be accepted as an honest and, to the extent that individual memory permitted, reliable witness.
Each of those witnesses conceded also the obvious point as to apparent deficiencies in shotcrete application in MC5B, as demonstrated in the diagram attached to Mr Jones' SI 173. Again, each was right to do so.
The witnesses were less willing to accept what Mr Williams put to them was the obvious inference following from what he said were deficiencies, appearing from the documents, manifested in the MCAA downdrive. Bearing in mind what I perceive to be the inadequacies in those records (or, more accurately, in the validity of any comparison of metres advanced to numbers of rockbolts used and quantities of shotcrete applied that may be made based on them), I do not see this as being a criticism of any of the witnesses.
I should say that Mr Bondin and Mr Wilson had suggested in their affidavits that some of the records might be inaccurate, specifically to the extent that they recorded the metres advanced. There was no hard evidence to support those suggestions. The impugned records included GSDs prepared by Mr Gilchrist, recording among other things the metres advanced (based on surveys undertaken daily). In the absence of hard evidence, I am not prepared to find that there was any significant inaccuracy in those records.
Mr Donaldson postulated that there were deficiencies in the rockbolt and shotcrete records. However, given that those records were maintained by TJH, and that TJH, in whose power it lay to explain or qualify them, did not do so, this suggestion does not rise above the level of speculation.
Mr Wilson referred to a survey which, he thought, showed that more rockbolts had been installed than appeared from the rockbolt records. If that survey were in evidence, I was not referred to it in the course of submissions. I do not feel inclined to undertake my own search through the huge electronic database that comprises the Court Book in this case for the purpose of seeing whether I can locate some document that might correspond to the "survey" that Mr Wilson described. One might think that, if there were such a document, he could have been taken to it in re-examination. There was no such re-examination.
It is apparent that Dr Beck had had regard to some survey reports (see for example T602.28). Again, however, those reports were not specifically identified. If (as may well be the case) they were among the documents provided to Dr Beck for the purpose of giving expert evidence, and referred to or exhibited in any of his reports (or annexed to any of them) they would have been admitted on the agreed basis that they demonstrated relevant assumptions of fact that he was asked to make, but did not prove the truth of those assumed facts.

[30]

The expert evidence

The experts who dealt with this in their reports were Dr Beck (for TJH) and Mr Kotze (for PSM).
Dr Beck dealt with this topic (and many others) in his report of 24 June 2015. He did so primarily by analysing "the records of actual ground support work undertaken by [TJH] in the intersection" (at page 10). Dr Beck then set out the history as it appeared to him from the documents that he reviewed (which included GSDs, and bolting and shotcrete reports). He said that although the history that he gave "is possibly incomplete, as any record may be, it gives sufficient evidence of ground support that was very likely installed in critical areas of the intersection" (again, page 10).
This section of Dr Beck's report was prepared in reply to modelling undertaken, and the subject of a report, by Dr Diederichs. Dr Diederichs was an expert to be called by PB. The disputes between TJH and PB were resolved before the expert evidence was taken. Dr Diederichs' reports were not read (nor were those of another expert to be called by PB, Professor Barla). Although Dr Diederichs and Professor Barla had participated in the conclave of experts and the joint report, the joint report was tendered and admitted on the basis that there would be notionally excluded from it the comments in it that were attributed to them.
Dr Beck returned to the topic at [7.1] of his report, where he dealt specifically with the topic of "adherence to the design". He noted that the design called for roof support comprising 4m rockbolts at 1m grids, followed by the application of 100mm of "fibrecrete" (as he called it; i.e., shotcrete), followed by arrays of 5m bolts at 1.5m grids, and a final 100mm layer of fibrecrete. Dr Beck concluded, in substance, that the support regime had been substantially adhered to in the MCAA/MC5B intersection, and westward into the downdrive.
Dr Beck's position was explained most clearly in the course of the concurrent evidence session.
The first point that Dr Beck made is the obvious one, made already, that a simple comparison of metres advanced with quantities of rockbolts and shotcrete used is not informative where the advance is, of necessity, a partial face advance. There is no need to spend more time over this. It should however be noted that, to the extent that this passage of cross-examination proceeded by reference to a drawing prepared by Mr Kotze, Dr Beck asserted that the drawing was not accurate in all respects.
When the cross-examination moved to the "actual" situation in the downdrive, Dr Beck said that the correct number of bolts had been installed (T599.1-.11): MR WILLIAMS: Taking into account whatever else you say you know, you would accept, wouldn't you, that over that period, 28 October to 31 October inclusive, there was a seriously deficient amount of shotcrete and rockbolts installed in that part of the works? DR BECK: Not taking into account what I actually know? So we have moved from your hypothetical now to the actual and if we take into account the actual and consider the surveys, then, in my view, one can form a view that they did install the bolts they were meant to install.
Mr Kotze was asked some questions about the matter. He said, and it is obvious, that in the downdrive, if one were to compare the metres advanced with the quantities of rockbolts and shotcrete used and assume a full face advance, there would be a serious deficiency. He was asked what would be the position if the advance were on a partial face. He did not express a view as to whether, in those circumstances, the quantities of rockbolts and shotcrete used might be relevant. In his view that was so because, as he said (T600.32-.33): That would prevent the design being applied because that was a deviation from the design.
I interpose at this stage to note that, since the downdrive could not be excavated full face in one pass of the header, I do not quite understand this answer. If Mr Kotze were suggesting that the design required partial advances of 1 metre on one side, being squared up immediately by an equivalent partial advance on the other, then it might make some sense. But this ignores the point made by Dr Beck, that until the design of the south-western corner of the intersection had been finalised, and the support installed (and neither of those things had happened at the relevant time), it was, at the very least, impractical to proceed by squaring up the downdrive as it headed west.
Returning to the question of alleged construction deficiencies by reference to quantities of rockbolts and shotcrete used, Mr Williams sought at T600-601 to tender a new report prepared by Dr Burman (after the conclave and joint report) that did express views on those matters. I rejected the tender.
Mr Williams in cross-examination of Dr Beck then returned to this question. I set out a lengthy but important passage of cross-examination covering both this and the question of finalisation of support at the southwest corner of the intersection (T602.27-605.19): DR BECK: Because you must understand that my assessments are based on actual surveys, not stylised drawings. Obviously I take an actual survey, showing face advance over any sort of summary that may have occurred at any particular time on a given day. So I consider the face shape, the wall position and the advance from survey to be the best estimate of where a tunnel was. I have not myself, for a very long time, computed the shotcrete quantities that would be required, given the particular way they advanced the face, and I haven't tallied the exact amounts of shotcrete that was installed on the particular days you said. We started on the 28th, which causes some difficulties. There is an apparent incongruity in the records between the 27th and the 28th that would make a large difference to the calculations, for example. So I'd be happy to go away and do those calculations, but I could only do them with reference to the actual survey. MR WILLIAMS: It is the case, isn't it, that you have no other record, other than the shift summaries and the shotcrete application summaries, that would indicate how much shotcrete had been applied over that period? DR BECK: That is correct, how much shotcrete was applied, that is correct. MR WILLIAMS: I want you to assume, if you would, that the calculation from those records is 8.6 cubic metres over the - DR BECK: For those two days, yes. MR WILLIAMS: That would be a seriously deficient amount of shotcrete for that advance, wouldn't it? DR BECK: The other part of that quotient is the surface area to which it has to be applied. So to compute how deficient or whether, indeed, it was deficient, one would have to compute the actual surface area to which it needed to be applied. MR WILLIAMS: You know that it is seriously deficient, don't you; you're just resisting giving me the obvious answer. DR BECK: I believe that there is a misunderstanding as to the completeness of the walls during those times and that could lead you to a vastly different calculation of the surface area which had to be covered. Now, if one were to presume that both the north and south walls in that area had to be completely covered during that period, and if that were, indeed, correct, then it would be seriously deficient. If those walls were not in their final position and, therefore, could not be fibrecreted, then one would need to compute whether the absence of completing that excavation to its final position, thereby leaving some rock in place, was equivalent support. MR WILLIAMS: Do you see yourself as an advocate for the plaintiff in this case? DR BECK: No, I don't, but I do see myself as an advocate for facts. In the course of the questions that I have been asked, I've had to review the actual survey, so I have had to reconstruct from original evidence the position of the walls and the position of the face. That was my original job in 2010, so I have always had a question as to certain things in relation to those particular walls. MR WILLIAMS: And you have never had a question about there being more than 8.6 cubic metres of shotcrete, have you? DR BECK: No, I'm not saying there was more than 8.6 cubic metres of shotcrete. I believe that is how much was sprayed. MR WILLIAMS: And based upon the assumptions that you have been given or the conclusions that you have drawn, you understand the position to be that the left-hand side of the down-drive was not shotcreted to the floor, is that right? DR BECK: Absolutely. MR WILLIAMS: And, therefore, there'd be no arch that would be formed by reason of shotcrete and bolting? DR BECK: My understanding is that there was no requirement to fibrecrete intermediate walls, so walls that were not in their final position. So as one turns that corner, you have to gradually chamfer that corner and one must excavate that wall in order to push it back and one can't simultaneously have complete fibrecrete and be excavating that wall. So since it is still at least a few metres from its final position, it is just physically not possible to have the fibrecrete on that wall yet. MR WILLIAMS: The design, as you understood it, for the ground support was an arch to be formed with shotcrete 200mm thick from wall to wall. DR BECK: That is the final design, that is correct. MR WILLIAMS: Yes, and there was no arch at all of the nature that the design called for by way of shotcrete from 28 October through to 31 October, was there? DR BECK: The corner was still being chamfered, so there could not have been. MR WILLIAMS: So the design intent of supporting this area by means of the shotcrete and bolting couldn't have been working, could it? DR BECK: It could not possibly have been because it wasn't yet excavated. MR WILLIAMS: It had been excavated for about 7.6 metres without any proper ground support, I suggest. DR BECK: Your suggestion is that there wasn't any proper ground support. My understanding is that the team - the designers, the constructors, the geotechnical team, were aware of the need to chamfer the corner. I haven't come across an explicit site instruction, other than a very generic plan view requesting that such a chamfer take place, that explains exactly how that corner is to be formed with that particular roadheader. So I don't know how the designer intended for that chamfer to be formed. I haven't seen what they indicated, how they wanted that to happen.
I should say at this point that although Dr Beck was eager (as indeed was Dr Burman) to ensure that his views were put before the Court in appropriate detail, I did not (contrary to the thrust of a question put in the course of this passage of cross-examination) regard him as an advocate for the cause of TJH (nor Dr Burman as an advocate for the cause of PSM).
Both Dr Beck and Dr Burman are very experienced, highly qualified and (so far as I can assess it) exceptionally capable in their field of expertise. Each struck me as someone who was capable of considering and, if appropriate, accepting, reasoned argument contrary to a position that he had stated. Each struck me as tenacious in the examination and analysis of his views. It does not follow, and I most definitely do not think, that either could be regarded as an "advocate". Each of them seemed to find the constraints of the "question and answer" process somewhat limiting. Certainly, each sought to escape from it where he thought it was appropriate, from a professional perspective, to do so. Again, I do not regard that in either case as a marker of advocacy.

[32]

Failure to comply with recommendations in the GSDs

This question is raised by para (c) and, in part, by para (f).
As to para (c), the focus of Mr Williams' submissions was on the MC5B backwall (which was also part of the northern wall of MCAA, through the intersection).
Mr Gilchrist's GSD of 17 October 2005 (see at [214] above) noted that MC5B had reached " ̴ 198 (backwall)". He wrote: Survey have marked up on shotcrete the start of the 5M bolt zone and appears that ̴ 3 rows of six bolts have been installed @ 1.5 m centres from ̴ CH185. …
Mr Gilchrist made recommendations as to the support system to be installed in the intersection and in the updrive and downdrive. His recommendation referred to drawing TW03-0059-1.
Mr Gilchrist's subsequent GSDs up until 27 October 2005 noted nothing of present relevance. His GSD of 27 October 2005 did. Among other things, he recommended that the MC5B backwall be trimmed and bolted: MC5B BACKWALL - once trimmed, spot bolts are marked up! Two rows 1.5 m longitudinal spacing from CH 482 down to left hand drive (BACK-DRIVE)
Clearly enough, Mr Gilchrist thus directed that, once the MC5B backwall was trimmed, it should be bolted in accordance with the pattern that he described.
Mr Bondin took note of Mr Gilchrist's direction. In SI 174, issued on 27 October 2005, Mr Bondin (dealing with the topic of "support for MCAA and MCBA and MC5B") noted, among other things: MC5B backwall Once the back wall has been trimmed, place spot bolts in marked locations
Mr Gilchrist's GSD of 28 October 2005 noted: - no bolts are installed in MC5B backwall yet, as am told this requires trimming
Mr Gilchrist's recommendations for MC5B included bolting the backwall in a pattern that he described. It may be noted that this GSD is the one that authorised deferral of the rockbolting of the left hand side of the MCAA downdrive (which Mr Gilchrist referred to as the "backdrive").
On 28 October 2005, the site diary records: MC5B - END WALL SPOT BOLT LOCATIONS HAVE BEEN SURVEY CHECKED AND ROCK DOES NOT NEED TRIMMING. PLEASE INSTALL THE SPOT BOLTS
Mr Gilchrist's next GSD, dated 31 October 2005, records no complaint as to the absence of bolting in the MC5B backwall.
In this context, although para (f) complains of failure to complete the excavation and support of MC5B before starting work on the MCAA updrive and downdrive, the simple fact is that Mr Gilchrist was well aware that this had happened. It is recorded in his GSDs. There is no suggestion in his GSDs that it was inappropriate, let alone a departure from design, for work to proceed in the way that it did.
That observation leads to what, to my mind, is the most salient feature of this aspect of PSM's complaints of defective work on the part of TJH. The obvious witness to give evidence of any such defective work was the author of the GSDs - Mr Gilchrist. Mr Gilchrist was not called to give evidence although, as I have said, PSM served a statement of evidence from him.
Thus, the Court is left to glean what it can from Mr Gilchrist's GSDs. In circumstances where the GSDs make no complaint of any failure to comply with any recommendation in them, the obvious inference is that Mr Gilchrist did not think that there was any, or at least any relevant, noncompliance. I draw that inference all the more comfortably bearing in mind the unexplained failure to call Mr Gilchrist.
In short, I am not satisfied that there has been proved any failure to comply with any relevant recommendation in any relevant GSD.

[33]

Failure to proceed by full face excavation

This is the issue raised by para (d). There are two aspects. One is whether the relevant design required a full face advance. The other is whether, in any event, and in the particular circumstances applying in the downdrive at the relevant time, good construction practice required a full face advance.
As I have said already, a decision on this paragraph is relevant not only in its own terms, but also in relation to some of the other paragraphs of issue 4.
Mr Williams did not point to anything that could properly (or even arguably) be described as a "design" that required excavation to proceed by way of a full face advance. Thus, on the evidence, there is no basis for concluding that the excavation in the MCAA downdrive by way of partial rather than full face advance represented a departure from design.
I am prepared to accept that, to the extent practicable at least, good construction practice would have required that TJH proceed by way of full face rather than partial face advance in the MCAA downdrive. However, as Dr Beck pointed out, there were practical difficulties, including that the support design for the left hand side of the downdrive at the point of the intersection had not been resolved.
Again, as it seems to me, the best evidence of what might have been good construction practice comes from Mr Gilchrist's GSDs. He was in the tunnels on virtually every working day. He was well and truly aware of the way in which excavation was proceeding. He did not note, in any GSD that is in evidence, that to proceed by way of partial face advance was poor construction practice, or otherwise undesirable. On the contrary, from time to time he made recommendations as to the support to be installed in the left hand wall of the downdrive, and as I have noted effectively authorised completion of that installation.
In the circumstances, I am not prepared to conclude that in the particular circumstances obtaining in the MCAA downdrive, the failure to proceed by way of full face advance represented poor construction practice.

[35]

Installation and tensioning of rockbolts

This relates to para (g).
The evidence on this comes from Mr Kotze (based on his initial investigations shortly after the collapse, and recorded in his WorkCover report) and Mr Wilson. Mr Kotze's evidence dealt with his inspections of rockbolts that had been recovered from the collapsed material. Mr Wilson's evidence dealt (on this issue) with the practice followed in the installation of rockbolts.
Before looking at that evidence, it should be noted that both Mr Kotze and Mr Peck agreed that rockbolts were unsuitable to be used as ground support for the conditions that prevailed. There was no evidence to the contrary, and I accept as correct this aspect of their evidence.
In Mr Kotze's view, rockbolts were inherently inappropriate as a support mechanism in the ground conditions prevailing at the intersection. He said at [20] of his WorkCover report: … the bedrock in and above the intersection is characterised by closely spaced low strength, vertically persistent fractures, that effectively divide the rock mass into a serious of semi-vertical columns. The orientation of the installed rock bolts, was often consistent with the orientation of the semi-vertical columns of rock that comprised potential failure masses. As dilation of the fractures occurred and the failure of roof developed propagating upwards along the semi-vertical fractures, the similarly oriented rock bolts were limited in their capacity to provide resistance to the failure mechanism.
Mr Peck said at [9.2] of his first report (26 March 2014) that the specified rockbolts were not appropriate for the conditions, having regard both to particular characteristics of those rockbolts and to the geological qualities of the prevailing rock mass. It is not necessary to go into detail.
Taking those matters into account, it is not surprising that (as Mr Wilson said) the installation of the rockbolts was difficult.
Mr Kotze inspected "a total of 59 remnant CT rockbolts… recovered from the roof collapse debris… ranging from virtually complete bolts through to various rockbolt fragments" (WorkCover report, page 22). He said, of those 59 bolts, that: 1. 39 had preserved threaded (mechanical) anchor ends, incorporating a nut fitting that could be wound so as to tension the bolts against the rock face at its inner or distal end; 2. 16 of the 39 "displayed wound nut positions that could have facilitated and maintained the expansion of the… anchor, thereby allowing end anchorage"; 3. the other 23 (of the 39) appeared to have been wound to an extent that end anchorage could not be achieved; and 4. of those 23 rockbolts which could not have achieved end anchorage, six were wound out to such an extent that grouting could not be applied, as required, between the rockbolt and the drill hole, so that in any event those rockbolts could not bond with the rock mass.
In my view, those observations must be regarded with very considerable caution. Of course, I do not doubt their factual accuracy. The question which requires careful consideration is, rather, what (if anything) they tell us about the installation of the rockbolts.
I should say that Mr Kotze refers consistently to "CT" rockbolts whereas the drawings (and Mr Peck) refer to "PB" rockbolts. Mr Peck explained that the PB rockbolt is a variant of the CT rockbolt, the principal variation being a longer thread for expansion of the anchor at the distal end.
Mr Wilson described the process of installation of rockbolt, in circumstances where a bolt could not be properly installed (because it "failed to tension") at [128] and following of his affidavit affirmed 26 July 2013. He said that such failures became more prevalent close to the dyke, because the ground was soft and the bolt would not hold. Where that happened, he said, the practice that he instructed tunnel crews to follow, was that the bolt should remain in the ground, and a second bolt should be installed next to it. He said that he observed that practice being followed.
Mr Wilson did say that some failed bolts were taken out of the roof. However, he said, if failed bolts could not be removed, they were bent over and covered with shotcrete.
Mr Wilson said, further, that if there were problems with rockbolt tensioning, he raised it at the 10 o'clock "toolbox" meeting. When that happened, he said, he was instructed simply to replace rockbolts that did not tension properly.
Mr Wilson gave oral evidence about the installation of rockbolts. He was cross-examined as to the requirement for "pull testing". It was put to him that this had never been done in MC5B or MCAA, nor in the intersection. He said that he would be surprised if this did not happen, but accepted that he had not seen it happen (T395.24-396.21).
Mr Wilson also gave oral evidence about work practices in relation to rockbolts that did not tension, or did not tension properly. He agreed in cross-examination that they would be left in place, and that a replacement bolt would be installed about 200mm distant (T425.28-.32). In re-examination, he said that when this happened, the failed bolts were not recorded in the rockbolt sheets (T450.16-.26).
There is one other piece of evidence that bears on this particular factual issue. After the collapse in the intersection, TJH received a report from PB dated 8 December 2005, dealing with the testing of rockbolts. Presumably, TJH had sought that report. PB said in substance that rockbolt testing to its satisfaction and that of its "onsite tunnel engineer" (perhaps, Dr Lloyd) had been performed in specified sections of the MC5A, MC5B, MC5C and MCBA tunnels. The chainages where that testing was carried out were specified in each case. The report stated: The location of the bolts tested was representative of the geology, as per PB's requirements. TJH reports that 6.1% of the bolts were tested, which complies with PB's requirements in approved Specification [the number was given]. All non-conformances reported by TJH to PB through their QA process, related to rockbolt testing, were reviewed and dispositioned [sic] by PB.
The Specification to which PB referred in that response required that for each design package, 10% of the first 100 rockbolts of each type or variant should be tested, and that thereafter 5% of the remaining rockbolts of each type or variant should be tested.
The four tunnels referred to in PB's advice were all part of the Marden Street Temporary Works Design Package.
As a matter of logic, it does not follow from Mr Wilson's evidence that he did not observe pull testing of the rockbolts in MC5B or MCAA (or in their intersection), that no such testing was carried out. Although Mr Wilson visited those tunnels two or three times a day, his duties required him to spend much time elsewhere. But even if one were to accept that no such testing was carried out in those locations, it does not follow that testing was not carried out in accordance with the Specification. And PB's report to which I have referred proves that appropriate testing was carried out.
The testing of rockbolts is really a side issue, but since it was raised in cross-examination, it was convenient to deal with it. However, the real importance of Mr Wilson's evidence is that it does not enable any inference, as to the quality of installation of rockbolts in and around the intersection, to be drawn from Mr Kotze's examination of the 59 bolts or bolt remnants that were recovered after the collapse.
There is no way of knowing how many of those 59 bolts were ones that TJH had categorised as bolts that had failed to tension, or to tension properly. There is no way of knowing how many of those 59 bolts had been categorised as not having been properly grouted. As a matter of logic, it might be thought that in a collapse of the kind that occurred, bolts that had not tensioned (at all, or properly), or had not been grouted (at all, or properly), might be more likely to dislodge in one piece than bolts which had bonded, mechanically or through the application of grouting, to the rock mass. But even if this be no more than impermissible speculation, the simple fact is that, because some failed bolts were left in situ, the 23 bolts that, in Mr Kotze's view, were unlikely to have anchored properly may well have been specimens of those failed bolts. If the bolts had been overwound without tensioning, it is likely that this would have been perceived by the crew installing them. It is thus likely that a replacement bolt would have been installed, as Mr Wilson said was done.
In short, when Mr Kotze's evidence on this topic is viewed against the background of what actually happened on site, the inference that can be drawn is that, as Mr Kotze said, there were a number of bolts that, obviously enough, had not tensioned (either properly or at all). It does not follow from Mr Kotze's evidence that the practice of replacing bolts which failed to tension, about which Mr Wilson gave evidence, was not followed. As I have said already, I accept Mr Wilson as a witness of truth who gave accurate evidence so far as his memory permitted.
The other matter to bear in mind in this context is Mr Peck's and Mr Kotze's uncontroverted evidence, that the geology was not suited to the installation of rockbolts. In those circumstances, it is not surprising that a higher number of rockbolts than "usual" might fail to tension, or to tension properly.
Looking at the evidence as a whole, I am not satisfied that TJH's work was unsatisfactory in relation to the actual installation of rockbolts (as opposed to the numbers installed). On the contrary, accepting as I do Mr Wilson's evidence, I find that when rockbolts could not be installed correctly, replacement rockbolts were installed.

[37]

Conclusions on the 4th and 5th issues

As to paras (a), (b), (d), (e) and (h) of issue 4, Mr Donaldson acknowledged, and I find, that there were failures of the kind referred to in each of those paragraphs.
In relation to paras (a), (b) and (in one respect) (c), I find, on the basis of the documentary records, the evidence of the lay witnesses and the evidence of the expert witness that I have dealt with at length above, that there were failures to comply with relevant SIs, and failures to comply with the cut one bolt one policy, to the extent that TJH excavated in the MCAA updrive without adhering to the cut one bolt one policy. Thus, there was a failure to adhere to SIs and GSDs that in effect specified the implementation of that policy. I do not find that there were any equivalent breaches in the MCAA downdrive. For the reasons I have given, the documentary evidence (and expert explanation of it) on which PSM relied does not prove the asserted failures.
Those failures in the updrive (to the extent they occurred) meant that TJH did not comply with the design constituted by Mr Gilchrist's GSD of 17 October 2005 and its attached drawing, as modified by Mr Bondin's SI 169 of the same date and its attached drawing.
However, in respect of para (b), the cut one bolt one policy was a construction practice or methodology, not an aspect of the design.
Further as to para (c), Mr Gilchrist at no time recorded, in any GSD that is in evidence, a complaint that a recommendation made by him in earlier GSDs had not been followed. I find that apart from the tangential matter referred to at [259] above relating to the cut one bolt one policy, there is no evidence of failure to comply with recommendations in GSDs.
As to para (d), it is the fact, and I find, that the excavation in the MCAA downdrive proceeded by way of partial face rather than full face excavation. However, that was not a failure to comply with any design that has been proved. It was an aspect of construction practice or methodology.
As to para (e), it is clear, and I find, that the thickness of shotcrete initially applied in MC5B (before the intersection) was inadequate. That was picked up by the checks ordered after the wedge failure occurred in the roof of the MCAA updrive, and was rectified.
There is no evidence that the shotcrete applied in the updrive was of inadequate thickness, to the extent that it was applied at all. Likewise, there is no evidence that the shotcrete applied in the downdrive was inadequate as to its thickness, to the extent that it was applied.
As to para (f), again there is no evidence in any of Mr Gilchrist's GSDs that are in evidence of complaint by him, or dissatisfaction on his part, as to the alleged failure the subject of this paragraph.
To the extent that para (f) relates to support to the MC5B backwall (i.e., part of the northern wall of MCAA), the best evidence comes from Mr Gilchrist's GSDs. They show that he was content for the installation of the 5m rockbolts to be deferred. See at [448] to [456] below.
As to para (g), the evidence does not support a finding that there was any failure to install and properly tension rockbolts. For the reasons I have given, whilst I accept the numerical and factual accuracy of Mr Kotze's observations based on his investigation, I do not regard them as demonstrating any deficiency in the installation and tensioning of rockbolts.
As to para (h), whilst the evidence does support a finding of failure to install the number and pattern of rockbolts required by the design in the updrive, it does not enable a similar finding to be made in respect of the downdrive.

[43]

The approach to design in the Marden Street Temporary Works

The proper starting point, in my view, is to be found in the Final Design Report prepared by PB relating to rock reinforcement. Clearly enough, PB was aware of the terms of that document, and so was TJH. PSM and URS must have been (or, at the very least, should have been) aware of it: apart from anything else, it was available on the project database, to which the various consultants and sub-consultants had access. In any event, I cannot understand how URS could have performed its obligations of review, or PSM its obligations identified at [174] to [176] above, if they had not considered, among many others, this document.
Although I have referred to this Final Design Report already, I shall set out, for ease of reference, what I perceive to be of present relevance.
Section 3 dealt with "Design Philosophy" and section 3.1 dealt with "Tunnel Support". The first paragraph of section 3.1 indicated that it applied to both temporary and permanent works.
The tunnel support philosophy was described as follows: The support systems comprise spot and pattern rock bolts / anchors, lattice girders, steel sets and shotcrete. The extent of the application of the temporary and permanent support system depends on the encountered rock types during construction, as classified by the tunnel geologist, and the construction sequences adopted, as determined by TJH Construction and the Designer. In areas of competent sandstone and siltstone rock possessing near-horizontal bedding planes and sub-vertical joint sets, rock bolts in combination with FRS will be used. The exact number of bolts and shotcrete thickness will vary depending on the class of the rock mass encountered and the span. The tunnel will have a flat roof profile. In areas of highly weathered rock mass, fault zones or altered ground (mainly in the Ashfield Shale), support will be provided using passive support techniques such as canopy tubes, spiling bars, sequential heading and bench methods, steel sets and lattice girders and fibre reinforced shotcrete (FRS). An arched tunnel crown is constructed.
Section 3.4 dealt with "Design Verification and Validation". The first matter addressed was the need to make checks on "insitu [sic] conditions": 3.4.1 Insitu [sic] conditions Checks of insitu conditions during the course of tunnelling operations are required to confirm design assumptions relative to actual rock behaviour. Some of the particular inspections include the following:

Ground classification including the geological and groundwater flow conditions;
Rock bolt / anchor pretension, and the capability of the mechanical anchorages;
Surface preparation of the rock prior to shotcreting;
Shotcrete and rock bolt / anchor bond;
The timing of support installation and
Monitored ground movement.

Section 4 dealt with "Tunnel Support Methodology". It noted that the design of the tunnel support was to involve a number of processes including: 1. studying similar large-span tunnels in Hawkesbury Sandstone or Ashfield Shale; 2. Voussoir beam analysis; and 3. numerical modelling of the "rock mass response".
A little later, the report set out a summary of the approach to support design: The support design is not based on providing a single universal design for the worst case condition but rather relies on insitu geotechnical observations and mapping to vary the support locally to address variation in conditions observed in the field and a range of support designed for these various condition.
Section 4.2.5 dealt with "Bolt Length". It stated, among other things, that: Based on analyses for worst geological conditions to be supported by rock reinforcement alone (Class III Sandstone and Class II Shale) and the experiences in different projects, a maximum roof sag of 15mm is assumed to be acceptable. Appendix E shows detailed worksheets showing the Voussoir beam analysis for sandstone and shale.
Three points should be made at this stage. The first is that "Class III Sandstone" can be disregarded, since this was an excavation in shale. The second point is that the reference to "Class II Shale" is unclear in the body of the report, but is explained in Appendix D as being shale classified no lower than G6. The third point is that the reference to "Appendix E" is erroneous and should be a reference to "Appendix D".
Appendix D makes it clear that a design which relied on the formation of a Voussoir beam, or which was said to be justified by Voussoir beam analysis, might be undertaken in ground classes G1 to G6, but not in ground classes G7 or lower. Indeed, Appendix D specifically states, among other things, that: The Voussoir-analysis is not applicable in classes G4, G5, G7 and G8. Passive support is provided for classes G5, G7 & G8. In class G4 an arched roof profile is adopted.
Although Appendix D does not specifically refer to ground class G9, it is inconceivable that a support method (or method of design analysis) that was clearly regarded as inadequate for shale of low or very low strength (G7 or G8) might have been considered to be appropriate for shale of extremely low strength (G9).
A Voussoir beam is a support method which relies on the rock mass above the excavation forming a natural beam to support the material above that beam (and, of course, the beam itself). Where the rock mass is fractured or otherwise broken up, rock bolting may assist in consolidating (my word) the rock mass, so as to facilitate the formation of a Voussoir beam, or so as to assist in Voussoir beam analysis.
Section 9 of the Design Report dealt with "Risk". Section 9.3 considered this topic in relation to "Tunnel Sections in Shale". What was there said made it perfectly clear that where the shale could not be considered "good", support was to be undertaken with structural steel ribs and shotcrete, or lattice girders and shotcrete: At portals and for the running tunnels and many ancillary or ramp tunnels, the crown sections lie within weathered Ashfield Shale. This rock is considerably less competent and has less strength than the Hawkesbury Sandstone. To compensate for this, the tunnel section is arched and supported with shotcrete and rock bolts where the shale is considered good, and with structural steel ribs and shotcrete, or lattice girders and shotcrete elsewhere. By using the arch section the ground loads are taken by the rock itself and supported by a relatively stiff support system. These sections will be excavated in staged excavations with limits of advance and heading size dictated by ground conditions.

[47]

The expert evidence

None of the experts who were called to give oral evidence thought that the design for the intersection and its surrounds was adequate. Mr Peck said, as I have noted at [352] above, that rockbolts were not appropriate in ground conditions of the kind actually encountered. Mr Peck said also that (in the MCAA tunnel), Voussoir beam analysis was inappropriate. Since that view was not controversial, and is plainly consistent with Appendix D to the Final Design Report, it is not necessary to go into Mr Peck's reasons.
I accept both those aspects of Mr Peck's evidence.
The view of the experts on the question of adequacy of the designs for the intersection is captured adequately in their response to question 4, set out in their joint report. That question 4 inquired whether the designs were appropriate having regard to what was known of the geological conditions or what would have been known had sufficient investigations been undertaken.
The joint report records the following: 4. The experts do not agree. (a) David Beck says no for the following reasons: (i) In this case, the ground quality, geometry (a wide intersection close to surface) and steep dipping dyke were all known to exist in the intersection as early as mid-August 2005. Poor ground also not only enveloped the excavation, but extended to surface, beyond the length of the longest rock bolts. The presence of electrical and water services and the slip road to the Pacific Highway also could have contributed to chimneying potential, by further promoting propagation of the failure, once it had commenced but these are not necessary for the initial failure to occur. The roof of the excavation was also not shaped to any meaningful extent to promote a stable arch. (ii) In addition there was a failure to analyse the capacity of the design as required to verify that the design was sufficient for the conditions. (b) Warren Peck agreed with David Beck but adds the following additional reason for saying no: (i) The design was not appropriate because its roof would have to rely on Voussoir Arches (a self supporting roof beam in jointed rock) for support which were unlikely to be stable in the Ashfield Shale due to the presence of the dyke, the low strength of the Ashfield Shale above RL 79 metres and the known existence of unfavourably-oriented discontinuities. Its span was to be 14 metres whilst the largest known span in the Ashfield Shale up to that time (2005) is believed to have been 5 metres. (c) Brian Burman says no for the following reasons: (i) There was no engineering design for the support system in the Final Design in terms of engineering calculations. That said, I accept that it is possible for a design to be based on experience and that the experience with the design of the MC5B tunnel was satisfactory, however, there was an apparent absence of consideration as to how that support system might be made fit for purpose for the particular circumstances of the intersection. The post collapse analyses prepared by the plaintiffs' experts on their face show that the design was not appropriate particularly in terms of its durability requirements. It was not sufficient to show, if that is what the plaintiffs' experts' models purport to show, that collapse was unlikely.
The experts who did not agree with those views were Dr Diederichs, Professor Barla and Dr Lay. They were not called. As I have said, it was common ground between the surviving parties, TJH and PSM, that their views should be notionally excluded from the joint report.
The experts' answers to question 5, 6 and 7 are also relevant, and I set out first those questions and then the answers (omitting the answers of those experts who were not called): Questions 5. What ground classification would a reasonably competent designer have assumed for the purposes of designing the tunnel support for the intersection of the tunnels? 6. Did the design of the intersection of the tunnels purport to contemplate the possibility of ground support classification of G9 and if not, what ground support classification did it purport to contemplate? 7. In the circumstances of the response of Q6 above to what extent was the observational method relevant to the design of MC5B/MCAA intersection and its implementation? Answers 5. The experts understand that the reference to ground classification is a reference to the LCT Classification System on drawing TS04-0052-0-0 and that the time at which the question is directed is the time when the design of the tunnel support for the intersection was prepared. (a) Warren Peck, David Beck, Brian Burman and Greg Kotze: G9 … 6. The experts do not agree on the answer to this question. (a) Brian Burman: (i) The design did purport to contemplate the possibility of ground support classifications of G9; (ii) The reason for that is that 200mm thickness shotcrete lining system which is consistent with G8 and G9. The statement by Dr Doug Maconachie (PB principal tunnel designer) at RFI TW03 019-TJH136 dated 18 November 2004; (iii) In September 2005 PSM advised PB that the dyke was expected to occur within the intersection. Damien Leis by email 29 September 2005 provided the support design which caters to the projected dyke at the intersection. Final FRS ensured 200mm shotcrete arch. … (d) David Beck…: (i) the final applied design does not purport to contemplate the possibility of ground support classifications of G9; (ii) it most closely matches ground classification of G7 or better. (iii) The reason for that view is that in every one of the tunnel support notes (eg, drawing PB DG TU TW03 004 350) ground classification G8 and G9 were always related to designs with steel sets. Ground classification G7 or better relied instead on shotcrete and rock bolts. The design for the intersection was closest to Mar VII which was for G4, G5 and G7 and showed 200mm of fibrecrete (FRS) but included 5m long bolts. 7. The experts are in partial agreement. (a) David Beck…: (i) G9 is the worst ground classification possible; (ii) That the observational method was relevant to the design of the MC5B/MCAA intersection; (b) The experts agree that: (i) Any geotechnical design assumption that affects the ability of the excavation to meet its design intent needs to be confirmed during construction. The observational design approach/method is employed for this purpose and is relevant to all geotechnical construction. (ii) In this project, an observational procedure for managing deviations from the assumed ground conditions or the ineffectiveness of support elements, or the non-compliance with design was meant to be employed. The relevant steps during construction, according to the WMS, were: 1. Examine ground as it is encountered during tunnelling; 2. Confirm/disconfirm whether the encountered ground conforms to expected conditions; 3. If it does conform, determine whether the GSC and excavation procedures should indeed be implemented. 4. Assess and evaluate the performance of the installed support and excavation procedure by means of appropriate observations and instrumentation; 5. If the ground is not as expected and is not suitably addressed by the design, then 1. the designers shall be notified; and 2. shall prepare a new design 6. Any modification of an existing design or preparation of a new design is to be instructed/authorised by the TJH design director or his/her delegated design manager allowing appropriate review times 7. If the installed support in any part of the tunnel does not conform, inform the area's project manager and witness rework for conformity; (iii) This process is consistent with the observational design approach: the purpose was to trigger design modification in the event a geotechnical assumption or performance of the support was different to what was expected. (c) Brian Burman, Warren Peck and Greg Kotze disagree that the observational method was relevant for the following reasons: (i) The design for the intersection that had been prepared as at 29 September 2005 had been frozen (email 29 September 2005 from Damien Leis to Sam Jones of TJH attaching PB DG TU TW03 0059-10-0 with a copy to Phil Lloyd); (ii) That design contemplated the worst case in terms of ground classification; (iii) There was no alternative design available; (iv) The extent of the observational method was limited to the direction of localised rock bolting.
It is convenient at this point to note that I do not agree with the view expressed by Dr Burman, Mr Peck and Mr Kotze that the design was frozen. I have already dealt with this, in relation to PSM's contractual obligations. However, when they say that "[t]here was no alternative design available", that is not literally correct.
It may be that they meant "available in the circumstances where the design had been frozen". If that is so, then no more need be said. If, however, the proposition were one intended to be understood at large, then, plainly, it is incorrect. As I have pointed out already, there were other designs prescribed for ground conditions worse than G7 in the Marden Street Temporary Works.

[48]

Para (a): did PB discharge its obligations?

On the evidence, the answer must be "no". PB's own design philosophy required it to provide a passive support design for the ground conditions that were actually encountered in the tunnel. There has been no explanation for, let alone justification of, the apparent abandonment (or disregard) of this aspect of the design philosophy.
At the time Mr Leis prepared the design set out in drawing TW03-0059-1 attached to his email of 29 September 2005, it was known that the dyke would reappear in MC5B and, likely, would traverse the intersection. It was also clear that the ground conditions in MC5B were not good.
In the nineteen GSDs that had been issued leading up to 29 September 2005, Mr Gilchrist classified the ground conditions as G7/G8 on numerous occasions, as G8 on five occasions, and as G7 on five occasions. There was no basis shown in those GSDs for any assumption that, from CH 168 in MC5B (the distance reached as recorded in Mr Gilchrist's GSD of 29 September 2005) up to and through the intersection, the conditions would improve consistently to G7 or better. On the contrary, the projected reappearance of the dyke, coupled with the known geology of the Ashfield Shale in the region, must have made it at least foreseeable (I would have said, very likely) that the ground conditions would be worse than G7 for almost the whole of the remaining section of tunnel.
Since PB withdrew from the hearing once it had resolved its differences with TJH, we do not know why it was that Mr Leis thought that the modified MAR VII design, as shown on his drawing TW03-0059-1, would be adequate in the conditions likely to be expected. Nor do we know why Mr Leis thought it appropriate to depart from the stipulations in the final design report and in drawing TW03-0041-5.
There is another complication with the design shown in drawing TW03-0059-1, which was only exposed (at least, for me) in the course of the concurrent evidence session. That drawing contemplated that the support for the crown of MC5B in the intersection strictly so called would comprise rockbolts of the stated length installed in the stated patterns, and a 200mm arch of shotcrete from floor to floor. However, it must have been obvious to everyone that the shotcrete arch would be destroyed once the tunnelers broke out into the updrive and the downdrive. Of necessity, that would require removal of the shotcrete on the relevant parts of the MC5B walls. What was left in the crown of that tunnel might have helped to prevent loose pieces of rock from falling out. It could not have offered structural support, for the obvious reason that the arch did not continue down to solid bedrock at the floor of the tunnel.
That is another criticism that Mr Peck made of the design for the intersection. He said, in substance, that it could never have worked as intended once the breakouts occurred. I agree. I add that, as was common ground, the support design for the south-western corner was never finalised.
All those matters, in my view, point to fundamental and serious deficiencies in PB's design.
In my view, it is plain that PB did not discharge its obligations to TJH in respect of the designs and in respect of its construction phase services.

[49]

Para (b): did PSM discharge its obligations to TJH?

PSM's suggested answer, "yes", depended critically on the views of its experts Dr Burman and Mr Kotze. There were in essence two strands to their reasoning on this question: 1. the design was frozen, so that whatever obligations PSM had in respect of monitoring and reporting on the adequacy of the design were effectively suspended; and 2. in any event, PSM was entitled to assume that the designer, PB, had done its job. It was not required to second-guess what PB had done.
To some extent, those strands are related. But, as it seems to me, they are not really matters that fall within the proper province of expert evidence (at least, of the type that Dr Burman and Mr Kotze are abundantly qualified to give). I accept that if I were to conclude (for example) that the design had been frozen, then Dr Burman and Mr Kotze might be able to give expert evidence as to what a qualified and competent engineer or engineering geologist, might have done in the circumstances that prevailed, in response to the observed rock conditions. And as will be seen, their evidence on this question may be relevant in connection with issue 13, which raises s 5O of the Civil Liability Act.
Mr Peck appeared to agree with the first strand of the reasoning of Dr Burman and Mr Kotze. He disagreed strongly with the second, as did Dr Beck.
In the light of my conclusions as to the extent and nature of PSM's contractual obligations, and my conclusion that the design was not "frozen" so as to suspend those obligations, the first reason that PSM suggested does not support the answer to para (b) for which PSM contends.
Once it is appreciated that there is no contractual answer, it seems to me to be plain that PSM did breach its obligations. There is no evidence at all to suggest that anyone from PSM - specifically, the SRME, Mr Clark - undertook any assessment of the suitability of the modified MAR VII support in the conditions that were both encountered and predicted to continue. In my view, the obvious conclusion, based on the absence of any documentary evidence of assessment and taking into account the unexplained failure to call Mr Clark or Mr Gilchrist, is that no such assessment was made. Nor can it be correct to say that PSM was entitled to rely on PB. The whole purpose of the observational approach to design, and of PSM's obligations relevant to that approach, was to require continual reassessment of the adequacy of PB's designs in the considerations actually encountered.
I should note that one of the issues contemplated by para (b) relates to suggested departures from the design on the part of TJH. In substance, I have dealt with this in connection with issue 5. However, there may be a separate question in respect of Mr Bondin's SI 169 and its accompanying drawing, and Mr Gilchrist's GSDs of 27 and 28 October 2005.
It is convenient to start with the design for the intersection, as Mr Kotze described it in his initial report (to WorkCover) at [11]: The excavation and ground support sequence adopted at the intersection was a modified Mar VII staged approach. (Figure 7). A face advance distance of 1.5m was prescribed, with the following roof support sequence:

The installation of 4m long rock bolts on a 1m x 1m grid pattern
An initial application of 100mm of shotcrete after face advance
Excavate remaining bench or benches as necessary
The installation of 5m long rock bolts on a 1.5m x 1.5m grid pattern
A final application of shotcrete to achieve a minimum total shotcrete thickness of 200mm (to form an arch from floor to floor). Such a sequence needs to be completed in its entirety, for full roof support to be achieved. Time delays within this sequence correspond with periods when the roof is not fully supported. At the time of the incident, the northern or right-hand side of the MCAA down drive from the intersection, had been advanced by more than the prescribed 1.5m. (Figure 1 and Reference 4). It is understood that the MCAA down drive had been partially installed on its right-hand side only with a number of 4m long rock bolts and with a single application of shotcrete of unknown thickness. (Figure 8). The partial excavation of the down drive on its' right-hand side only, prevented the establishment of a shotcrete arch across the full drive width. See Plates 12 and 15. The excavation and ground support sequence had not been completed. The excavation roof was therefore less than full supported. The limited support in place at the time could not accommodate the geological conditions encountered in the roof and roof collapse occurred.

Mr Kotze referred to his figure 7. That was drawing TW03-0043-5, which described, among other things, the full face construction sequence MAR VII design. There is no present need to go to the other references given by Mr Kotze in this paragraph of his report.
When one goes to Mr Bondin's SI 169 and its drawing, that is precisely the support system that he described. Although Mr Kotze referred to the partial face advance in the downdrive, that was not a feature of the drawing attached to Mr Bondin's SI 169. That drawing appears to describe the excavation in both the updrive and the downdrive taking place by way of full face advance.
Mr Gilchrist's GSD of 27 October 2005 commented on the excavation in and adjacent to the intersection. In relation to what Mr Gilchrist called the "backdrive" (that is, the downdrive), his recommendations stated: - Dyke affected ground - 4m long bolts @ 1 m centres + handle face plates - 100mm shotcrete (min) in cycle - 5m to be installed later - total 200mm shotcrete thickness (min) to continue
Thus, in this GSD, Mr Gilchrist specifically authorised deferring the installation of the 5m rockbolts.
On 28 October 2005, Mr Gilchrist's GSD made essentially the same recommendations for the downdrive: - 4m longrock bolts @ 1m centres + handle face plates - 100mm shotcrete in cycle - 5m long rockbolts to be installed later @ 1.5m grid - total 200mm shotcrete thickness to continue.
In short, in these GSDs, Mr Gilchrist accepted that installation of the 5m long rockbolts at 1.5 m centres, which were an integral part of the modified MAR VII design on which Mr Kotze had commented at [11] of his report, could be deferred.
To the extent that PSM now wishes to make out a case that the failure to install the 5m rockbolts at 1.5m centres in the downdrive was a departure from the design or a construction deficiency, the simple fact is that it was a departure, or an approach to construction, recognised and expressly authorised by its representative Mr Gilchrist. It may be noted that on the inspection of 27 October, Mr Gilchrist was accompanied by Dr Lloyd; and on the inspection of 28 October, he was accompanied by Mr Jones.
There is another aspect of those two GSDs which should be mentioned at this point. It was I think part of PSM's case, as to construction deficiencies, that the failure to trim the MC5B backwall (that is, the northern extremity of that tunnel which in turn formed part of the eastern wall of MCAA), and the failure to install spot bolts as recommended by Mr Gilchrist, represented either a departure from design or a departure from good construction practice. TJH's response to this argument was that: 1. it had not been shown that the failure to install those spot bolts had any causal relationship with the collapse that occurred on 1 November 2005; and 2. Mr Gilchrist had known that the spot bolts had not been installed, and had acquiesced in the delay in installing them.
The first point is not of present relevance, but is undoubtedly correct. I do not think that any of the experts demonstrated a causal connection between the failure to install those spot bolts in the back or side wall (depending on which way one looks at it) and the roof collapse.

[50]

Para (c): did URS discharge its obligations of verification?

Mr Williams' submissions addressed this only in respect of the drawing that was verified, TW03-0059-1. Mr Williams relied on evidence given by Mr Peck. Mr Peck said that: 1. a site inspection had not been done (presumably by URS) of the intersection, despite the known risk of faults and dykes; 2. the strength of the rock layers had not been established, although it was known (according to Mr Peck) that the strength of the roof strata was about half that of the tunnel walls; 3. there were no stability computations done for the intersection (presumably, by PB); and 4. URS should have discovered that no computations had been done and that no investigative bore holes had been drilled.
Mr Peck then speculated as to "an ongoing discussion" between URS and PB, apparently on the basis of his interpretation of the ways in which various drawings had been marked up and were (or appeared to have been) exchanged. That does not seem to go anywhere, in circumstances where it has been shown that the drawing in question was verified.
Mr Peck accepted that URS "was not the checking or proof engineer… and, hence, was not obligated to check the accuracy of the various tunnel design calculations". However, he said, URS "was responsible for ascertaining that appropriate tunnel design calculations had been done by a competent tunnel design engineer".
The real problem with this aspect of PSM's case is that the paragraphs of Mr Peck's report (dated 18 June 2015) on which it relies appear to reflect his understanding based on various documents that had been given to him for the purpose of preparing his opinions. It has not been proved that there had been no site investigation. It has not been proved that the strengths of the various rock layers had not been established. It has not been proved that no bore holes had been drilled. And, crucially, it has not been proved that there were no stability computations performed.
Mr Peck was not involved in the design or review process. His statements of those matters must reflect either his understanding of primary material that is not itself in evidence (or, if it is in evidence, was not referred to in the course of submissions), or assumptions made by him based on his review of the primary material. Either way, it seems to me, those statements are not probative of the facts asserted by them.
It may be - I do not know - that among the material that URS might have relied upon had its case not been resolved were primary documents that might have established some or all of those conclusions or assumptions. If there were such materials, they were not referred to in the course of submissions.
Mr Peck accepted that URS was not the checking or proof engineer, and thus was not obliged to check the accuracy of tunnel design calculations. That is not a conclusion that is within his province as an expert. It is however, if I may say so, an accurate summary of URS' contractual obligations (more accurately, of the limitations on those obligations). The case against URS must therefore be that it did not use reasonable care to satisfy itself that the design for the intersection had been prepared in an adequate or professionally appropriate way. The absence of evidence as to the conclusions or assumptions made by Mr Peck, on which PSM relied, is fatal to that case.
I conclude that it has not been shown that URS breached its contractual obligations of verification.

[51]

Para (d): proper discharge of PB's and PSM's obligations in respect of GSDs

I am not sure how this issue arises in respect of PB. Mr Williams submitted that the answer should be "yes" in respect of both PB and PSM. Mr Donaldson's submissions did not address the position of PB. I proceed on the basis that the issue need be answered only in respect of PSM.
The experts considered this topic in their joint report. In their answer to question 10 (which concerned the adequacy of the GSDs produced by PSM given the information and data available to it), three of the experts (Dr Burman, Mr Kotze and Mr Peck) concluded that the GSDs were reasonable. Mr Peck did however qualify his concurrence in that agreement by saying that to the extent the GSDs referred to "anything less than G8 in the vicinity of the dyke", they were "optimistic".
Dr Beck agreed that the mapping in the GSDs was reasonable. However, he said, the determinations were not reasonable because, in his view, the ground conditions in and around the intersection should have been classified consistently as G9. He referred to the geological characteristics of the ground, by reference to the ground strength classification system to which I have referred above.
Dr Burman, Mr Kotze and Mr Peck disagreed with Dr Beck on this, in part because they thought that the classification system was insufficiently precise, such that "the conditions in the intersection could have been assessed as any of G7, G8 or G9". Further, they said, Dr Beck's opinion disregarded some aspects of the classification system "and specifically rock strength".
The matter was explored in cross-examination. It think it is fair to say that the experts remained of the differing views expressed by them in the joint report.
In circumstances where the question of classification is necessarily one of skilled judgment, requiring the application of specialised skill and knowledge to observed ground conditions and the assessment of those conditions against the classification scheme, I think it appropriate to proceed in accordance with the majority view. I am fortified in that approach by the consideration that (as does not seem to be controversial) there was some lack of precision in the classification system itself.
In any event, so it seems to me, it is very difficult indeed to divorce analysis and criticism of the classifications adopted from the knowledge of what happened afterwards. I think that this has influenced the approach taken by Dr Beck. To put the matter less obliquely, I do think that there is a significant element of hindsight in Dr Beck's reasoning on this aspect of the case.
I conclude that PSM did perform its classification duties with appropriate professional skill.

[54]

The legislation and its effect

Section 5O of the Civil Liability Act provides a defence of "widely accepted competent professional practice". As Giles JA (with whom Ipp and Basten JJA agreed) observed in Dobler v Halverson (2007) 70 NSWLR 151 at [59], the intention of the legislature was to reintroduce, into the law governing professional negligence, a modified version of the principle often described as the "Bolam" principle, after Bolam v Friern Hospital Management Committee [1957] 1 WLR 582.
Section 5O reads as follows: 5O Standard of care for professionals (1) A person practising a profession ("a professional" ) does not incur a liability in negligence arising from the provision of a professional service if it is established that the professional acted in a manner that (at the time the service was provided) was widely accepted in Australia by peer professional opinion as competent professional practice. (2) However, peer professional opinion cannot be relied on for the purposes of this section if the court considers that the opinion is irrational. (3) The fact that there are differing peer professional opinions widely accepted in Australia concerning a matter does not prevent any one or more (or all) of those opinions being relied on for the purposes of this section. (4) Peer professional opinion does not have to be universally accepted to be considered widely accepted.
Giles JA described the operation of s 5O in Dobler at [59] to [61]. His Honour observed at [59] that the effect of s 5O was that if the professional's conduct was of a kind that would be widely accepted as competent by peer professional opinion then, as long as that peer professional opinion was rational, it set the standard of care. Thus, as his Honour said at [60], s 5O provides a defence. As his Honour noted at [61], s5O did not operate only to define the standard of care.
I set out those paragraphs: 59. Section 5O was amongst the tort law reforms consequent on the Review of Law of Negligence, Final Report (Canberra, Commonwealth of Australia, September 2002) (the Review). It was intended to introduce a modified Bolam principle. Its importance does not lie so much in questions of onus of proof as in who determines the standard of care. Commonly, as in the present case, there will be expert evidence called by the plaintiff to the effect that the defendant's conduct fell short of acceptable professional practice and expert evidence called by the defendant that it did not; the expert evidence may or may not recognise that the opposing professional practice is one which has some currency. Apart from s 5O, the Court would determine the standard of care, guided by the evidence of acceptable professional practice. It would not be obliged to hold against the plaintiff if the defendant's conduct accorded with professional practice regarded as acceptable by some although not by others. Section 5O has the effect that, if the defendant's conduct accorded with professional practice regarded as acceptable by some (more fully, if he "acted in a manner that … was widely accepted … by peer professional opinion as competent professional practice"), then subject to rationality that professional practice sets the standard of care 60. In this sense, s 5O provides a defence. The plaintiff will usually call his expert evidence to the effect that the defendant's conduct fell short of acceptable professional practice, and will invite the Court to determine the standard of care in accordance with that evidence. He will not be concerned to identify and negate a different professional practice favourable to the defendant, and s 5O does not require that he do so. The defendant has the interest in calling expert evidence to establish that he acted according to professional practice widely accepted by peer professional opinion, which if accepted will (subject to rationality) mean that he escapes liability. 61. It follows that I do not accept the appellant's submission that s 5O did not provide a defence but defined the content of the duty of care owed by the appellant to Kurt, with the onus on the respondents to prove that the manner in which he acted was not widely accepted by peer professional opinion as competent professional practice. Section 5O may end up operating so as to determine the defendant's standard of care, but the standard of care will be that determined by the Court with guidance from evidence of acceptable professional practice unless it is established (in practice, by the defendant) that the defendant acted according to professional practice widely accepted by (rational) peer professional opinion. To require the plaintiff to establish the negative would significantly distort the language of s 5O(1), and would not be consistent with the reference in s 5O(2) to reliance on peer professional opinion for the purposes of the section - the plaintiff does not rely on it in order to negate a liability in negligence.

[56]

Decision

Dr Burman said merely that, on the basis of the documentation given to him, PSM did act "in the professional manner expected of ordinarily competent professionals providing the services for which they were contracted". However, as Dr Burman made clear, that view was inextricably linked to his view of the "limited and closely defined involvement" that, he thought, PSM had in the project. Further, and as Mr Donaldson submitted, Dr Burman did not address the specific requirements of the relevant contracts.
Mr Kotze also touched on this topic, but only in relation to the GSDs. Since I have concluded that PSM did provide its services appropriately in respect of ground classification, it is not necessary to look at Mr Kotze's evidence.
Dr Burman's expression of opinion on this point is, to put it mildly, conclusory. To the extent that he descends into detail, he does so more by seeking to rebut the view that Dr Beck and Mr Peck had of PSM's contractual obligations.
In my view, the question raised by s 5O cannot be considered in a vacuum. It can only be considered, and the widely accepted peer professional opinion can only be assessed, by reference to the specific obligations that the professional undertakes pursuant to the contract of retainer. I accept that there will be circumstances where the s 5O question arises otherwise than in the context of a contract of retainer, but since I am dealing with a very specific contract for very clearly defined services, the more general situation of obligations imposed by the common law may be put to one side.
It is easy to see how s 5O operates where the professional undertakes no more than the usual and proper obligation to exercise due care and skill in the performance of duties of design, inspection, supervision, or whatever else is the subject of the particular retainer. In the present case, the obligations that PSM undertook were very carefully designed to reflect the particular demands of this complex project. It may be that peer professional opinion could be relevant in the context with which I am concerned. It is not necessary to decide that point, because in my view, Dr Burman's opinion is not capable of demonstrating what might be competent professional practice, widely accepted by peer professional opinion in Australia, relevant to the performance of the specific and detailed obligations undertaken by PSM.
As I have noted, Dr Burman addresses the matter at a high level of abstraction. His analysis (to the extent that any is provided) depends critically on his view that the design for the intersection was frozen in September or October 2005. For the reasons I have given, I do not accept that analysis.
In those circumstances, it seems to me, Dr Burman's evidence goes nowhere near making good the s 5O defence.
As I have noted, s 5O is relevant to PB as well. There is no evidence that suggests that PB, in the relevant respects, did act in a manner that would be widely accepted as competent by peer professional practice in Australia. On the contrary, the common view of the experts who gave evidence (that is to say, those who were called by TJH or PSM) was to the effect that PB did not perform its obligations with appropriate professional care and skill.
Issue 13 should be answered "no", to the extent that it arises in respect of PB and PSM.

[58]

The expert evidence

Question 15, considered by the experts, asked: If the design of the intersection had been implemented, would the failure have occurred?
The answers given, in so far as they refer to those of the experts who ended up being called and who gave evidence in the concurrent session, were as follows: 15. The experts do not agree. (a) David Beck says the collapse would have occurred had the design been implemented for the reasons set out in answer to question above other than question 9 as well as the following: (i) In the actual conditions (G9, or possibly G8), a failure would have still occurred and at around the same time and to the same extent. This is based on the mechanism of failure, the known conditions which included a number of risk factors for chimneying as described above and the poor fit of the ground support to the conditions as well as David Beck's own numerical analysis. (ii) In better conditions (G7 or better) the design and or as installed support may or may not have prevented the failure but the FOS and durability would have been too low for the design intent. (b) Brian Burman says: (i) it is likely that the collapse wold [sic] have occurred had the design been implemented; and (ii) agrees with (a) (ii) from David Beck: … (d) Warren Peck says that the collapse would have occurred had the design been implemented because the faulted roof was unsupportable with rock bolts and shotcrete. Steel arches were needed to support the intersection's roof. (e) Greg Kotze does not offer an opinion in answer this question as he was not involved with the design and has not been involved in the computer modelling of the intersection.
Dr Beck carried out computer modelling intended to simulate excavation in the MCAA downdrive. His modelling assumed that the relevant design was that specified in drawing TW03-0059-1 (Mr Leis' design of 28 September 2005). Dr Beck modelled, in the alternative, a half face advance in the MCAA downdrive and a full face advance. He concluded, in substance, that it was more likely than not that the failure that did in fact occur would have occurred, at about the same place and in the same manner as it actually occurred, regardless of the suggested construction deficiencies.
Neither Dr Burman nor Mr Kotze carried out any modelling. They did not in terms criticize the modelling undertaken by Dr Beck, although of course the thrust of their evidence was, as I have said, that the alleged construction deficiencies were causally significant.
Mr Williams submitted that Dr Burman's agreement that "it is likely that the collapse would have occurred had the design been implemented" should not be taken as an acceptance of the proposition that the particular collapse would have happened, in the way that it did, had the design been followed. It might be thought that this proposition would have carried more weight had it come from Dr Burman. Regardless, because Dr Burman's view was founded on the asserted construction deficiencies in the MCAA downdrive, and because I have concluded that those deficiencies have not been proved, Mr Williams' reinterpretation of Dr Burman's apparent assent (as recorded in 15(b)(i) of the joint report) goes nowhere.

[60]

Causation - PSM

Mr Williams submitted that any hypothetical breach of duty on the part of PSM would have had no causal effect. That was so, he submitted, because Dr Lloyd (of PB) was a member of the Geotechnical team. Mr Williams submitted, in substance, that if Mr Clark or Mr Gilchrist had realised that the design was inadequate for the conditions being encountered, they would have taken this up with Dr Lloyd. Dr Lloyd in turn would have taken the matter up with Mr Leis. Mr Leis, presumably, would have confirmed the sufficiency of his design.
That argument falls down at a number of points. First, there was no evidence from Mr Clark or Mr Gilchrist. Thus, there is no evidence of what their actions might have been had they turned their mind to the apparent incompatibility of Mr Leis' design with the conditions actually being encountered in the intersection.
Second, although Dr Lloyd was an employee of PB, he was on secondment to the Geotechnical team. So far as the evidence goes, he was not a member of PB's Design team.
Third, there is simply no reason for assuming that Mr Leis, or Dr Maconochie, or some other responsible employee of PM, would have merely confirmed the adequacy of the design had concerns been raised. After all, the whole point of the observational approach to design was that there should be constant feedback, from those observing the ground conditions as work progressed to those responsible for those producing the design.
Fourth (and this really flows from the third point), the argument seems to me to overlook the essential commercial purpose underlying this aspect of PSM's obligations. That purpose was to ensure that the suitability of the designs was monitored continuously, so as to ensure (to the extent that proper professional performance of the parties' contractual obligations could do) that the project met its design requirements, including the one hundred year life span.
I add that Messrs Wilson and Wille gave evidence that if PSM had raised concerns as to the continuing suitability of Mr Leis' design, and recommended changing (for example, to steel sets), they would have followed that advice. That evidence was not challenged. I accept it.

[63]

Issue 18: damages; apportionment

As between TJH and PSM, damages are agreed at $20.95m, before interest. TJH claims interest. It is difficult to see why it should not have interest on its loss. Mr Williams did not address this point.
As to apportionment, those between whom the loss ought be apportioned are PB and PSM (bearing in mind my conclusion that TJH did not cause or contribute to the loss).
As the High Court pointed out in Podrebersek v Australian Iron and Steel Pty Ltd (1985) 59 ALJR 492 at 494, the exercise of apportioning responsibility for loss between two or more wrongdoers has two elements. The first is a comparison of the culpability of each wrongdoer - the extent to which it departed from the requisite standard of care. The other element looks at the relative importance of each party's acts in causing the damage complained of; their "causal potency" (Zanner v Zanner (2010) 79 NSWLR 702 at [102]). The exercise of apportionment involves a consideration of the whole of the relevant conduct of each party.
In the present case, it seems to me, each of PB and PSM departed in a very significant way from the standard imposed upon it by its contract with TJH. In PB's case, the support design that it produced for the intersection and surrounding areas did not reflect the design philosophy that PB itself had propounded. As to PSM, there has been a wholesale failure to assess the ongoing suitability of that design in the light of ground conditions actually reported, and to communicate the outcome of that assessment to PB.
In terms of causal potency, each failure seems to me to be an independent and effective cause of the loss that followed. Had PB adhered to its design philosophy and produced a design utilising passive support such as steel sets, then (so far as the evidence goes) the collapse would not have occurred. Had PSM assessed the design in the light of the ground conditions encountered, it should have raised with PB the adequacy of the design. Had PSM done so, it is likely that PB would have revised the support design to ensure that it was adequate.
However, even if each breach of duty were an effective cause, it does not follow that their contributions were equally potent. Assessment of causal potency requires more than an analysis of causation: see at [511] above.
In my view, looking at each breach in its full context, primary responsibility should be attributed to PB. PB had the responsibility of producing designs that were appropriate to the circumstances. The design that it produced was not appropriate, even at the time it was produced, having regard to: 1. PB's own design philosophy; 2. the known ground conditions in MC5B when the design was produced, and the likelihood that those ground conditions were, at best, unlikely to improve; and 3. the inherent unsuitability of rockbolts in the ground conditions that were known, and expected to continue (see at [350] to [352] above).
However, the causal potency of PSM's breach of its obligations was not minimal. It was obliged to monitor the design in the light of ground conditions. It was obliged to consider whether the design was appropriate. If PSM, having done those things, considered that the design might not be appropriate, it was obliged to communicate with PB. There is no evidence that PSM ever considered the ongoing suitability of the design, having regard to ground conditions that were encountered.
The observational approach to design was an essential feature of the project. It was intended to ensure that the support design chosen at any given point was and remained suitable having regard to the ground conditions actually encountered. It was, in a very real sense, intended to guard against what happened in this case: namely, the ongoing utilisation of a support design that was inappropriate to the conditions in which it was applied.
In my view, accepting the primary responsibility of PB, the failure of PSM to perform its contractual obligations was, nonetheless, significant.
It is difficult to convert impressionistic views of comparative responsibility (in the broad sense) to precise figures. Any attempt to do so gives a misleading semblance of certainty and precision to what is very much a matter of judgment and balance. The outcome is unlikely to be capable of detailed rational explanation. Acknowledging as I have just done the inherent frailty of the process, my conclusion is that responsibility should be assigned 2/3 to PB and 1/3 to PSM.

[68]

Decision - the Construction Risks Policy

I do not agree that PSM was not a subcontractor. Clearly, in my view, it was. Both the derivative obligations of TJH under the Project Deed and their direct obligations under the D&C Deed included design obligations. If, as Mr Donaldson submitted, PSM's obligations under its Consultancy Agreement were, or included, design obligations then, to that extent at least, TJH must be taken to have subcontracted part of their design obligations to PSM.
In my view, the definition of "Insured" in the Construction Risks Policy makes it clear that consultants such as PSM may be "sub-contractors". That follows necessarily from the words of cl 4.2. Cleary enough, architects and consultants will be insured as "sub-contractors", so long as their non-design activities are performed on site, or their design obligations (if any) are performed off site. There can be no doubt that PSM would fall within the generic description of "consultant".
However, I do agree that the relevant obligations of PSM were design obligations. That seems to me to follow inevitably from the observational approach to design. The design was not something fixed immutably at the commencement of the project. It was something to be assessed and, if necessary, modified, as the project progressed. PSM's obligations under its Consultancy Agreement, as I have identified them earlier, gave it a role in that ongoing process of design.
First, PSM was required to assess the suitability of the support system specified in the light of conditions actually encountered. And second, PSM was required to specify the support system to be adopted, in the light of ground conditions actually encountered. Of necessity, part at least of those obligations had to be performed on site.
In Vosten, Ryan J (who gave the judgment of the Full Court) considered whether the specification of the method of anchorage of the luffing rope of a crane constituted "design". The evidence was that the decision, how to anchor the luffing rope, was one made on site once the crane had been rigged. His Honour held that the selection of a particular method of anchorage was "a decision on the design of the method of anchorage".
In my view, that analysis applies in the present case. An obligation to decide whether a specified support design is appropriate, and should continue to be used, is an integral part of the design of the support to be used. Inded, that is so a fortiori bearing in mind the observational approach to design. And likewise, the specification of a support system to be used is an integral element of the design of that support system.
Had PSM made some attempt to comply with its obligations, and negligently specified an inadequate method of support, it would not be entitled to indemnity under the policy. I do not understand how it can be placed in a better position, simply because it made no attempt to perform those duties. The duties, for breach of which it is being sued, are properly characterised as design duties. From the perspective of the definition of "Insured", it matters not whether the complaint is that the duties were performed negligently, or that they were not performed at all.
Thus, in my view, PSM is not an "Insured" under the Construction Risks Policy. Clause 15 has no application. There is no bar to recovery, insofar as the losses claimed have been the subject of indemnity under the Construction Risks Policy.

[70]

Submissions - the Legal Liability Policy

Mr Williams submitted that TJH had been obliged to put the Legal Liability Policy in place for itself and, among others, PSM. That obligation arose, Mr Williams submitted, under cl 22.4 of the Project Deed and under cl 22.4 of the D&C deed.
Mr Donaldson submitted that neither the Project Deed nor the D&C Deed was relevant. PSM was not a party to either of those deeds, and cl 5 of the Legal Liability Policy focused attention on the contract between the relevant insured (in this case, TJH) and the subcontractor (in this case, PSM). It followed, Mr Donaldson submitted, that even if PSM were to be regarded as a subcontractor, it was not a subcontractor in respect of whom, by the subcontract between it and TJH, the latter were required to effect insurance.
Mr Williams relied, in particular, on the D&C Deed. He noted, correctly, that cl 22.4 imposed on TJH an obligation to effect third party liability insurance, and submitted, again correctly, that the Legal Liability Policy had been effected in pursuance of that obligation. Mr Williams submitted that by cl 22.5 of the D&C Deed, the policy was required to insure, among others, anyone having "an insurable interest under the Project Documents…" and was required to include a cross-liability / waiver of subrogation clause.
Mr Williams submitted that PSM had an insurable interest under a Project Document because it had been "retained by TJH in connection with the obligations of TJH pursuant to the D&C Deed and the Project Deed", both of which were "Project Documents". Mr Williams' submissions did not identify his client's Consultancy Agreement as a "Project Document". It would constitute a Project Document if the parties to the Project Deed (i.e., the RTA, LCTC and the Trustee) had agreed it to be one. Mr Williams did not point to any such agreement.
In any event, Mr Donaldson submitted, the Project Deed did not directly impose any obligation of any kind on TJH, because TJH was not a party to it. Mr Donaldson noted, correctly, that cl 22.4 of the Project Deed imposed obligations on the Trustee and LCTC (who, with the RTA, were the only parties to that deed).

[71]

Decision - the Legal Liability Policy

The starting point must be that as TJH was not a party to the Project Deed, cl 22.4 of that deed could impose no obligations on TJH. Thus, whatever relevance the Project Deed may have in the process of construction, it is not a relevant source of TJH's insurance obligations.
TJH was of course a party to the D&C Deed. Clause 22.4(a)(i) of that deed required TJH to effect and maintain third party liability insurance. And cl 22.5(vii) required that policy to include a cross liability and waiver of subrogation clause. However, that obligation only applied to those who were insured under the policy taken out in pursuance of cl 22.4(a)(i).
In my view, when cl 22.5(a)(iv) refers to those who "have an insurable interest under the Project Document", it is to be read as referring to those who are parties to, and accept obligations under, one or more of those Project Documents. Thus, in my view, cl 22.5(a)(iv) of the D&C Deed imposed no obligation on TJH to effect third party liability insurance for the benefit of its consultants, including PSM, unless their contracts with TJH were agreed to be "Project Documents".
If that conclusion were wrong, TJH would be in breach of its obligations owed to the other parties to the D&C Deed. That breach would sound in damages at the suit of one or more of those parties. But would not be a breach actionable at the suit of PSM, because PSM was not a party to the D&C Deed. Mr Williams did not submit that any of the other parties to the D&C Deed held the benefit of cl 22.4 on trust for PSM, or that PSM was entitled to sue on cl 22.4 by application of the principles explained in Trident General Insurance Co Ltd v McNiece Bros Pty Ltd (1988) 165 CLR 107.
In any event, the question is not what the position of PSM might have been under some hypothetical policy that, on its view, should have been effected. It is, rather, what is the position of PSM under the policy that was effected. And that policy is clear. By cl 5, which qualifies the general definition of the "Insured", a subcontractor will only have the benefit of the policy where the terms of the contract made between that subcontractor and (in this case) TJH required the latter to effect such insurance for the benefit of the former. Mr Williams pointed to no provision of the Consultancy Agreement between TJH and PSM which required TJH to effect such insurance for the benefit of PSM.
I conclude that PSM is not an insured under the Legal Liability Policy. It is not entitled to be indemnified under that policy, and cl 14 has no operation.

At a glance

Court

Decision date

Before

Source

Judgment (74 paragraphs)

Parties

Legislation Cited (4)

Cases Cited (5)

Cited By (2)