Scottish Court of Session Decisions

 

OUTER HOUSE, COURT OF SESSION

[2016] CSOH 177

 

CA162/12

OPINION OF LORD WOOLMAN

In the cause

SSE GENERATION LIMITED

Pursuer

against

HOCHTIEF SOLUTIONS AG & ANOTHER

Defender

Pursuer:  Moynihan QC, S Smith QC, Barne, Delibegović-Broome & P O’Brien;  CMS Cameron McKenna

Defender:  Clark QC, Richardson & A Sutherland;  Clyde & Co

21 December 2016

Contents

1. Introduction

   Summary

   Issues

   Preliminary Matters

2. The Principal Persons

   SSE

   Hochtief

   Donaldson

   Pöyry

   Andritz

3. Hydroelectric Tunnels

   History

   Design & Rock Classification

   Methods of Construction

   Lined & Unlined Tunnels

   Previous Tunnel Collapses

4. The Glendoe Scheme

   Geography

   Main Features

    Headrace Tunnel

    Tailrace Tunnel

    Turbine

   Other Aspects

    Head Loss

    Aqueduct and Access Tunnels

    General Operating Centre

    Watering Up and Dewatering

    Measurements

5. Award of the Contract

   Tender Procedure

6. The Construction Period

   Hochtief arrives on Site

   Tunnel Boring Machine

   The TBM Drive

   Task of the Engineering Geologist

   Four Excavation Classes

   Rock Support Methodology

   Mapping Procedure

   Geological Database

   Signing the REC sheets

   What Classes were Predicted and Found?

   Conagleann Fault Zone

7. From Watering up to Takeover

8. SSE Operates the Scheme

   Takeover

   O & M Manuals

   National Grid Tests

   Monitoring

   Glendoe’s Role within the SSE Fleet

   Odd Readings

9. The Collapse & Its Aftermath

   Detection

   Initial Strategy

   SSE Informs Hochtief

   Tunnel Inspections

10. Relations Cool

   Attempts to Reach a Consensus

   Conference Call on 25 September

   Meeting at Edinburgh Airport on 5 October

   SSE Meets HSE on 12 October

   Two Further Meetings on 20 and 21 October

   Essen Design Workshop on 29 October

   Hochtief’s Proposals

   SSE Issues an Ultimatum

11. Specialist Investigations

   Jacobs

   British Geological Survey

   Donaldson

12. The Recovery Project

   Commercial Pressures on SSE

   SSE Selects BAM

   Choice of Option E

   BAM Mobilises to Site

   Scope of the Remedial Works

   Contract Changes

13. Why did the Tunnel Collapse?

   The Short Answer

   A Longer Answer

14. Witnesses & the Burden of Proof

15. Was there a Defect at Takeover?

   Scope of the Dispute

   Was The Scheme Contrary to the Works Information?

   Was The Scheme Contrary to the Design?

16. Reasonable Skill and Care

   What did the parties agree about classification?

   Standard of Mapping

   Joint Inspections in 2008

   Three Miscellaneous Issues

    (a) Did the REC Sheets Address Long Term Hazards?

    (b) Was Hochtief’s Classification System Deficient?

    (c) Would Probe Drilling Have Made a Difference?

   Conclusion on Reasonable Care

17. Should Hochtief have Returned to Site?

18. Contributory Negligence

19. Causation

20. The Costs of the Recovery Project

   Introduction

   The Position of the Parties

   General Principle

   Choice of Option E

   Costs Monitoring

   Expert Evidence on Costs

   Health and Safety

   Betterment

21. Individual Heads of Claim

   Bypass Tunnel

   Downstream Access Tunnel

   Concrete Invert

   Three Concrete Plugs

   Plant & Equipment

   Secondary ‘Defects’

   SSE Management Time

   Progress, Delays & Productivity

   Adjudication Award, Insurance & Low Availability Damages

22. The Counterclaim

23. Observations on the Proof

24. Conclusions

25. Appendix

 

1. Introduction

Summary

[1]        Glendoe hydroelectric scheme was constructed between 2006 and 2008. It was the biggest such scheme to be built in Scotland for many years. It failed eight months after take over. A collapse had occurred in a main tunnel. Rock material had fallen from a breach in the crown over several months, leading to a blockage.

[2]        The owner was anxious to recover its asset as soon as possible. It was losing substantial amounts of revenue while the scheme was out of commission. It entered into discussions with the contractor about remedial works. They were, however, unable to reach agreement. The main sticking point was money. Each party said that the other should bear the initial liability for the cost of the works. Attempts to broker a compromise were unsuccessful.

[3]        Four months passed without a resolution. The owner then issued an ultimatum requiring the contractor to return to site to carry out the recovery project. The contractor declined to do so. In consequence, the owner instructed another engineering concern to undertake the remedial works. The works took much longer and cost far more than expected. The scheme did not begin to generate electricity again until August 2012.

[4]        This litigation arises out of the collapse. The owner seeks to recover about £130 million. It maintains that various provisions in the contract imposed obligations on the contractor to undertake the remedial works. The contractor disputes liability. It contends that: (a) it completed the works in accordance with the contract, (b) the owner assumed the risk of the collapse at takeover, (c) the scope and cost of the recovery project is unreasonable, and (d) it was entitled to carry out the remedial works.

 

Issues

[5]        There are several key questions in this case. Why did the collapse occur? Could it have been foreseen? Is the contractor to blame? Is the owner also partly to blame? Were the remedial works reasonable and necessary? The contract itself distinguishes between employer’s risk events and contractor’s risk events. It also prescribes the circumstances where the contractor is obliged to carry out the remedial works.

[6]        At the core of the dispute lies a deceptively simple issue. It relates to the construction of the tunnel. Did the contractor misclassify the rock and in consequence fail to install the correct support? The owner submits that the answer is Yes. The contractor takes the opposite position.  It maintains that it exercised reasonable skill and care in the design and construction of the tunnel and is therefore not liable.

 

Preliminary Matters

[7]        At an earlier stage, the contractor argued that the joint insurance policy governed the matter and that the owner was precluded from bringing this action. Following a debate I issued a decision in favour of the owner: [2015] CSOH 92. The contractor undertook not to reclaim that issue prior to the proof. In closing submissions I learned that the owner has made a claim under the policy, which is the subject of arbitration. If it received an award that would have to be taken into account in respect of any damages it received in this action.

[8]        I fixed the proof to take place over two court terms. At the end of the first term, the parties made a joint application to set aside a third term to deal with quantum. I refused that request.

[9]        It is important to keep in mind the distinction between the collapse and the blocked section of the tunnel. The location, length and scale of the blocked section are known. The same parameters in respect of the collapse remain unknown.

[10]      To avoid verbiage, I have rounded up most sums and measurements.

 

2. The Principal Persons

SSE

[11]      I shall refer to the Scottish & Southern Electricity group of companies as ‘SSE’. SSE owns and operates a fleet of 80 hydroelectric power stations. Ten lie in the Great Glen area close to Glendoe. The SSE fleet contains a total of 214 tunnels with a combined length of 320km. Despite its substantial tunnelling experience, SSE took a limited role during the construction of Glendoe. Its project manager, Mr Neil Sandilands, said that was because “the process was being carried out by one of the world’s top contractors and supervised by one of the world’s top consultants.”

 

Hochtief

[12]      Hochtief Solutions AG is a long-established engineering contractor with its headquarters in Germany. It has constructed rail, road and hydroelectric tunnels throughout the world. It set up an internal joint venture chaired by Bert Hoekstra to deliver the Glendoe scheme. It comprised various Hochtief divisions, including Hochtief UK Construction Ltd. I shall refer to the joint venture as ‘Hochtief’.

Jacobs

[13]      SSE appointed Jacobs Gibb Ltd and Jacobs UK Ltd (together ‘Jacobs’) to act as its engineering geology consultant at Glendoe. Jacobs had an important role in relation to the design of both the original scheme and the recovery project. During the construction of the original scheme, it also had a supervisory function. Mr Sandilands described Jacobs as being SSE’s “eyes and ears in the tunnel”.

 

Donaldson

[14]      SSE engaged Donaldson Associates Limited (‘Donaldson’) to provide technical advice on the recovery project. It also appointed Professor Andrew Sloan of Donaldson to act as its expert in this case.

 

Pöyry

[15]      Hochtief retained Pöyry Energy JV (‘Pöyry’), a Finnish company, as its sub-contractor for the design of the civil works. Mr John Ager of Pöyry regularly visited the tunnels to monitor the works and liaise with Hochtief.

 

Andritz

[16]      Hochtief also engaged Andritz, an Austrian company, to be its mechanical and electrical sub-contractor. In the early 2000s, SSE refurbished 25 of its power stations. Andritz was involved in that project, which standardised the electronic signals sent out by each plant.  It therefore had some knowledge of SSE’s operations.

 

 
3. Hydroelectric Tunnels

History

[17]      The science of tunnelling dates back to antiquity. Its modern history is linked to Scandinavia. After the First World War there was a shortage of steel for industry. Norwegian engineers discovered that water can be taken through unlined tunnels, provided that there is enough rock cover to counteract the water pressure inside the tunnel. The unlined tunnel concept was further developed in the 1950s as it afforded substantial savings in cost and time. Norway is now completely dependent on hydropower and has several thousand kilometres of unlined tunnels.

 

Design & Rock Classification

[18]      There are risks to a tunnel at each stage of its life: excavation, watering up, operation and dewatering. It is crucial to ensure that the correct level of support is installed within the tunnel. Otherwise it may collapse. There are several recognised classification systems. One is the ‘Q-system’, introduced in 1974. It is set out in Engineering Classification of Rock Masses for the Design of Tunnel Support, published by the Norwegian Geotechnical Institute, which has gone through several editions.

[19]      SSE envisaged that the Q-system would be used at Glendoe. It classifies rock mass quality by adding together certain numerical values. Subsequently SSE accepted Hochtief’s proposal to use an ‘observational’ system, which relies on the skills and experience of the engineering geologists to assess the rock mass in the tunnel.

 

Methods of Construction

[20]      The traditional method of tunnel construction is by means of ‘drill and blast’. It involves packing explosives into drilled holes. After each explosion the resultant debris is removed. A newer method involves the use of a tunnel boring machine (‘TBM’). It provides a tunnel with a smoother surface that may require less support. A TBM can, however, cause delay and expense if it becomes stuck in poor rock conditions.

 

Lined & Unlined Tunnels

[21]      Most railway and road tunnels have a continuous concrete or steel lining that goes around the whole perimeter. By contrast unlined tunnels use the ‘Roman Arch’ principle. The rockmass supports itself by developing an arch of competent rock directly around the opening. Despite the name, most unlined tunnels require some support in sections of weak rock and also close to the turbine. The distinction between lined & unlined tunnels is therefore not always clearcut.

[22]      Shotcrete (sprayed concrete) provides a level of support below that provided by concrete or steel. It provides a hard, durable cover that prevents the eroding effect of water. The bottom of a tunnel is known as the ‘invert’ and the roof as the ‘crown’.

 

Previous Tunnel Collapses

[23]      Tunnel collapses have occurred around the world. SSE itself experienced the collapse of an unlined tunnel at North Lochay in 1990. The tunnelling experts who gave evidence in this case referred to these prior collapses. In some instances, they had either been involved in the investigations or advised on the remedial works. All were familiar with the literature on the subject.

[24]      I derive little benefit from these prior collapses, however, because each one has to be studied in the light of its own circumstances. Two witnesses, Professor Stille and Dr Palmström, illustrate the problem. They are the co-authors of a standard text book on rock engineering that has gone through several editions. In this case, however, they hold radically different views. Professor Stille places reliance on a 1987 collapse at the Holmens Paper Mill in Sweden. Dr Palmström refers to a collapse at the Lower Vinstra hydroelectric scheme in Norway. I do not regard it as profitable to try to tease out the points of similarity and difference with regard to these other schemes. The focus must be on Glendoe.

 

4. The Glendoe Scheme

Geography

[25]      Glendoe lies in the Monadhliath Mountains near the southern tip of Loch Ness. It offers several advantages for a hydroelectric scheme. It is in a good water catchment area. Its high elevation of 600m gives it the potential to generate large amounts of electricity. The rock is generally of good quality for tunnelling.

[26]      But Glendoe also has disadvantages. The winters are harsh and the remote location makes access difficult. In addition, the Conagleann Fault Zone (‘CFZ’) poses a risk to tunnel integrity. The CFZ is a large tear in the earth’s crust. It was probably caused by an earthquake that happened tens of millions of years ago. While it is apparent on the surface, the only means of knowing how the individual fault strands align and intersect with each other underground is to drill boreholes.

 

Main Features

[27]      The Glendoe scheme is configured as follows. Water from a reservoir flows down a tunnel to a turbine deep underground. From there it flows through another tunnel into Loch Ness. The combined length of both tunnels is 8.2km (5 miles). They each have a diameter of 5m, roughly the same size as that of a London Underground tunnel. Hochtief used a TBM to excavate one continuous tunnel. After its completion, Hochtief inserted a concrete plug into the tunnel near the turbine. That resulted in the tunnel being divided in two. The headrace tunnel (6.2km) conveys water from the reservoir to the turbine. The tailrace tunnel (1.9km) conveys water from the turbine to Loch Ness. The parties agreed that the tunnels should have a design life of 75 years.

 

Headrace Tunnel

[28]      The headrace tunnel (‘HRT’) is very steep for much of its length. As the water travels down at high speed, it creates high pressure within the tunnel. To avoid the water bursting out, there must be at least 250m of rock above the tunnel. Although most of the HRT is unlined, it does have (i) a steel lined section near the power station, (ii) a section of reinforced concrete to resist dewatering problems, and (iii) support in the form of shotcrete and rockbolts at various areas.

 

Tailrace Tunnel

[29]      The tailrace tunnel (‘TRT’) is mainly level. As it is unpressurised, the water flows freely and does not run against the crown. Because the TRT has a less arduous life than the HRT, erosion damage is less likely.

 

Turbine

[30]      When the water arrives at the power station from the HRT, six needle valves direct it into buckets on the ‘Pelton’ turbine runner wheel. The degree of opening of the needle valves determines the power output. A rock trap prevents damage to the runner wheel from any loose material coming down the HRT.

 

Other Aspects

Head Loss

[31]      The ‘net head’ of a hydroelectric scheme is the pressure available to drive the turbine. It is calculated by subtracting the head loss from the gross head. The gross head represents the difference in height between the elevation of the turbine and the reservoir water level. When water flows through a tunnel there are friction or ‘head’ losses. Glendoe was designed to have a very low head loss of 0.4%. The net head at Glendoe was designed to be between 600m and 606m. 

 

Aqueduct and Access Tunnels

[32]      A tree of aqueduct tunnels conduct water into the reservoir. There is also an access tunnel to the power station for people, vehicles and equipment. Hochtief excavated these other tunnels by means of drill and blast.

 

General Operating Centre

[33]      SSE controls all its power stations from the General Operation Centre (‘GOC’) in Perth. The engineers at the GOC co-ordinate the fleet to match supply with demand. To do so they look at reservoir levels, rainfall forecasts and National Grid requirements. In carrying out their task they have regard to electronic signals and alarms from each station. In 2009, there were no alarms for needle valve opening, net head, or head loss.

 

Watering Up and Dewatering

[34]      Watering up and dewatering an unlined tunnel are both risky procedures. They may expose weaknesses that have not been identified during construction. In particular, the changes in pressure can trigger rockfalls. SSE had intended to carry out the first dewatering in May 2010 and then institute a six year dewatering cycle.

 

Measurements

[35]      Tunnels are measured in ‘chainages’. The term derives from the old practice of surveyors using fixed lengths of chain to measure ground. The HRT intersected the CFZ between chainages 2040 and 2210.

 

5. Award of the Contract

Tender Procedure

[36]      Following a pre-qualification procedure, SSE conducted a two stage tender process. In early 2004 it invited contractors to tender on the basis of a reference design, which involved several assumptions: (i) excavation by drill and blast, (ii) rock classification using the Q‑system, (iii) five rock support classes, and (iv) a continuous lining in the HRT.

[37]      SSE told contractors that Jacobs expected 97.5% of the excavation to be in ‘good’ rock, but added that there was a paucity of information about the ground conditions. It outlined the further investigations suggested by Jacobs. SSE asked tenderers to specify any further ones that they wished undertaken. It envisaged that some of these would take place during construction. The parties knew that exploratory drillings are not always reliable. Complex fault zones are sometimes only discovered during the course of the works.

[38]      Hochtief submitted its tender on 7 April 2004 setting out its rates and tunnelling method. It stated that: “During construction, the support requirement will be decided at the face, based on actual conditions”. It agreed with Jacobs’ proposed investigations.

[39]   In advance of submitting its tender, however, Hochtief had asked SSE to drill a deep borehole at ‘intake 15’ to find out more about the geology of the CFZ. SSE was unable to fulfil that request. It would have required the largest drilling rig in the UK. At that stage there was no access road to allow such a rig to be transported by lorry. No civilian helicopter could lift a rig of that size.

[40]      In August 2004 Jacobs produced a Ground Reference Conditions Report for the benefit of tenderers. It summarised the geological information then available. It referred in particular to borehole PT1, which had been drilled earlier that summer to discover the rock mass conditions where the CFZ intersected the tunnel. It indicated that the geology was relatively straightforward. Investigations after the collapse established, however, that PT1 missed the main area of the CFZ. Dr Smith of the British Geological Survey (‘BGS’) said that it misled the parties into thinking that there were good tunnelling conditions.

[41]      By October 2004 only two tenderers remained: Hochtief and Morgan Est. Each held meetings with SSE and Jacobs to discuss their technical proposals. Hochtief submitted a revised tender in November 2004. It stated that it intended to use a TBM and that it expected to leave 60% of the headrace tunnel unlined “without compromising the required service life”. It also intended to employ a classification system with four rather than five classes of support. Class  I represented good rock conditions. Class  IV represented very poor rock conditions.

[42]      In August 2005 SSE designated Hochtief as the preferred bidder. After further negotiations the parties signed a contract on 22 and 28 December 2005 (‘the contract’). It was based on Option A of the New Engineering Contract 2 (‘NEC’). The contract price was £126 million with a £2 million bonus for completion by 31 December 2008.

[43]      SSE had envisaged three design stages: concept, preliminary, and detailed. When the parties executed the contract, the first two stages had been completed. The necessary material was contained in the works information. SSE approved Pöyry’s detailed design before construction began. That included the Headrace TBM Excavation Rock Support Methodology (‘RSM’).  

 

6. The Construction Period

Hochtief arrives on Site

[44]      Hochtief established itself on site in early 2006. It constructed accommodation for its workforce, as well as setting up offices, vehicle workshops and stores. From the outset it developed a good working relationship with SSE, which endured throughout the construction period. Jonathan Bird of Hochtief, who has been in the tunnelling industry for over twenty years, said that SSE was the best employer with whom he had worked. The parties held weekly and monthly meetings to discuss technical issues, which they mainly resolved by agreement. Representatives of Jacobs, Pöyry and Andritz also participated in the discussions where appropriate.

 

Tunnel Boring Machine

[45]      The TBM was an enormous contraption, some 150m long. Its parts were delivered to site in the middle of 2006. It took three months to assemble the TBM. The local schoolchildren named it ‘Eliza Jane’.

[46]      The cutterhead at the front of the TBM had 34 individual cutters. They chiselled the tunnel face under high pressure. A conveyor belt attached to the tunnel roof removed the broken rock. The operator controlled the TBM from a cabin 20m behind the cutterhead. He set the cutter rotation speed, the thrust pressure and the gripper pressure. The grippers extended from the sides of the TBM to the tunnel walls. They pushed it forward in a ratchet type movement. The TBM operator had to take great care in the steep part of the HRT to ensure that the TBM did not fall backwards.

[47]      A probe drilling rig sat just behind the cutterhead. It could be used to explore the rock ahead for the presence of water and gas. Because the conditions were so dry, Hochtief never carried out forward probing.

[48]      A TBM is a dangerous place to work, particularly at the cutterhead. Hochtief employed an experienced crew at Glendoe. They worked under the direction of the TBM Manager. The crew for each shift comprised a shift foreman, an operator, two rock bolters, two shotcreters, several tunnelling labourers, a mechanic and an electrician. The TBM crew installed the first line of support (bolts and mesh) from close behind the cutterhead as soon as the rock was exposed. They generally sprayed the shotcrete from further back.

[49]      Multi-purpose vehicles with traction units at each end transported individuals and materials to and from the TBM. They had different ‘cassettes’ for passengers and equipment. 

 

The TBM Drive

[50]      The TBM began to excavate the tunnel in late September 2006. It advanced on average 24m in each twenty four hour period. It completed the TRT section in early February 2007. By late summer, it had travelled 6km. The rock conditions were very dry and there had been few sections of weak rock.

 

Task of the Engineering Geologist

[51]      In classifying rock an engineering geologist engages in a process of interpretation. He looks for any features that might pose a risk to tunnel stability. In particular he assesses the orientation and aperture of any faults in the rockmass to determine its strength.

[52]      Hochtief employed two engineering geologists on site. David Taylor was principally responsible for the HRT and the TRT. He did not give evidence. Rolf Wilhelm was responsible for other elements of the scheme, including the aqueduct tunnels. After he left the Glendoe project Wilhelm joined a firm of engineering consultants, Lahymeyer. He came back after the collapse and participated in some of the investigations. He said that he was offering independent views as well as giving factual evidence about the construction period.

 

Four Excavation Classes

[53]      The parties agreed that there should be four rock excavation classes at Glendoe. Class  I required minimal support. Class  II involved rock bolts and shotcrete around part of the crown. Class  III was full circumference shotcrete. Class  IV prescribed increased shotcrete, together with steel arches. These were set out on drawing D201l. The engineering geologist could also specify additional support measures within a particular class.
[54]      Pöyry provided six guidelines to assist in determining the correct class:

a)   Determine risk situations for which the rock support should be designed.

b)   Design a range of typical sections (one for each excavation class) covering foreseeable risk situations.

c)   Determine forecast for intensity of risk situations along tunnel length and hence for distribution of excavation classes along the tunnel.

d)   Determine actual support requirements during the tunnel advance based on actual conditions encountered, particularly with reference to the identified risk situations.

e)   Adapt rock support based on actual rock conditions.

f)    Observe the behaviour of the rock behind the face in order to judge the adequacy of the initial support and any need for additional rock support and/or final lining (such as shotcrete lining to prevent erosion of weak rock layers).

 

Rock Support Methodology

[55]      The RSM also contained two statements about the rock classification. First, paragraph 6.1.5 stated: “Erodible zones will have to be shotcreted irrespective of the associated rock class”. On 27 October 2006 Jacobs asked Hochtief to clarify what was meant by “erodible”. Hochtief did not reply.

[56]      Second, Table 11 of the tunnel support  methodology refers to the hazard of “erosion of erodible rock during operation” of the HRT. The countermeasure is “Application of shotcrete if not already covered/‘protected’ by steel rib support.” I shall return to the question of whether these statements are to be read in conjunction with paras (e) and (f) of the Pöyry guidelines and in particular whether there were mandatory requirement imposed on Hochtief.

 

Mapping Procedure

[57]      Before the commencement of the TBM drive, the parties agreed that there should be joint rock classification using the observational method. That is clear, for example, from the minutes of a meeting on 2 February 2006: “Both SSE and Hochtief will employ geologists on site to continuously agree rock classification”. Similarly the minute of a meeting on 24 May records that support decisions should “be made/confirmed on site”. 

[58]      Hochtief considered that it would be useful to devise a practical document to assist in this exercise. Accordingly, Taylor prepared a draft rock excavation classification sheet (‘REC sheet’). He revised it in the light of comments from his Hochtief colleagues and Jacobs. After SSE approved the REC sheet, it was used without query during the execution of the works.

[59]      The front of the REC sheet looks like a questionnaire. It lists various rock features with a tick box alongside. Taylor normally completed the sheet. He had (a) to decide which boxes to tick, (b) to determine the overall rock excavation class, (c) to prescribe any additional rock support, and (d) to add any comments. There was no direct correlation between the number of ticks and the overall rock excavation class. Nor were the features ranked in order of importance.

[60]      On the reverse of the REC sheet Taylor indicated the excavation class covered by the sheet, and identified the rock conditions that lay immediately ahead.

[61]      Taylor completed both sides of the REC sheet each time he was in the tunnel. According to Zimmerman he did so after discussion with Jacobs’ personnel. They jointly agreed the classification in the tunnel. Before leaving the tunnel, Taylor left a copy of the reverse of the REC sheet for the TBM crew.

[62]      Taylor also completed fold-out mapping sheets on which he also recorded his observations of the tunnel walls, including the type of rock and any faults. The fold-out sheets each covered a 25m section of tunnel. By contrast each REC sheet only covered a length with the same rock classification. It could therefore vary in length from 1m to 100m.

 

Geological Database

[63]      In terms of the RSM, Hochtief created a geological database for the purposes of “payment and design compliance”. When Taylor returned to his site office after visiting the tunnel, he updated the database and sent an email to Pöyry and Jacobs in the following terms:

“The latest rock excavation classification assessment and geological mapping can be found via the following link … Please advise if the classification or support specification is at variance with your observations or design assumptions.”

 

[64]      There was no evidence that anyone sought changes or otherwise queried Taylor’s decisions.

 

Signing the REC sheets

[65]      Hochtief, Jacobs and Pöyry all signed the REC sheets. Derek Williams of Jacobs signed them as “approved” on behalf of SSE, rather than the weaker terms of “as seen” or “acknowledged”. As he signed them many weeks later, he had ample opportunity to cross‑check Taylor’s classification against the findings of the Q‑system.

[66]      Sandilands, whom I found to be a candid and helpful witness, accepted that the RECs played an important role during construction, that Williams’ signature bound SSE, that the geologists discussed the appropriate level of support, and that “such a system would need to exist”.

 

What Classes were Predicted and Found?

[67]      Hochtief found the tunnel to be very dry and the rock mass conditions much better than predicted. That is shown in the following table.

   

	Expected	As Built
Class I	58.9%	84.3%
Class II	23.1%	15.4%
Class III	14.3%	0.3%
Class IV	3.7%	0%

  

 

Conagleann Fault Zone

[68]      Hochtief expected to encounter poor rock mass conditions in the CFZ. When the TBM was approaching that area, it stored extra supplies of support on the machine. It also sought the removal of the probe rig, as it impeded the rock bolters in carrying out their task. Jacobs agreed to this request, but only on being assured that it could be re-fitted within eight hours.

[69]      The TBM passed through the CFZ between 19 and 29 August. For four of those days, however, it was stationary for planned maintenance.  On 24 August Hochtief stated that it did not intend to probe ahead. SSE did not challenge that approach.

[70]      Taylor sent emails to his colleagues to warn them about the potential hazard. On 23 August he advised the TBM crew to take “extreme care” when walking through that area of the tunnel. On 28 August, he stated “Now in anticipated fault zone so expect deterioration at any time.”

[71]      The TBM crew found no signs of poor rock conditions in the CFZ. Mr Spiers, SSE’s resident engineer for tunnels, prepared the minute of the weekly progress meeting on 9 September. It records that the conditions were class 1 throughout: “The TBM has now passed through the predicted zone of the Conagleann Fault, but indications of its existence were imperceptible.”

 

7. From Watering up to Takeover

[72]      Prior to watering up, the parties conducted several ‘metre by metre’ inspections of the tunnel. Their aim was to detect any potential problems and to decide whether any sections required additional support.

[73]      The first inspection took place on 21 & 22 January 2008. Afterwards, Taylor proposed reclassifying some sections from class I to class II. A further inspection took place on 27 February to look at these sections more closely, which was attended by representatives of Hochtief, Jacobs and Pöyry. Subsequently, Pöyry prepared a design statement recommending the application of strips of shotcrete and mesh (a ‘band aid’ solution) to certain areas of erodible rock.
[74]      Professor Broch inspected the HRT on 11 March and reported:

“My general impression was that the rock mass conditions in the tunnel are very good. It is one of the driest tunnels that I have ever inspected ... No major weakness zones were observed ... The weakness zones described in the Pöyry report are all small, and as far as I could observe, none of these may cause serious collapses. I thus regard the support measures recommended in the report as being more than good enough for a tunnel that is basically designed and built as unlined.”

 

[75]      In court, Professor Broch said that he inspected practically all the locations described by Pöyry and walked the parts of the tunnel in between. He took samples to check for swelling clay minerals, which can be a sign of weakness. Tests at the Norwegian Technical University, however, did not indicate any potential problems.

[76]      Before watering up, there were further joint inspections of the entire HRT to ensure that the works complied with the contract requirements and the Pöyry Design Statement. Mr Speirs expressed a concern about the lower side wall. Hochtief and Jacobs reassured him that if erosion occurred it would not adversely affect the stability of the tunnel.

[77]      Otherwise no one expressed any concern about the overall stability of the tunnel. At the handover inspection on 25 and 26 October, SSE noticed some minor items lying in the invert. Hochtief removed them the following night. SSE intended to review the tunnel at the first maintenance inspection, which was scheduled for May 2010. It would be normal for some further support to be installed at that stage.

[78]      In the spring of 2008 Hochtief tried to find out more about the CFZ. It drilled a borehole at Intake 15 – the location it had identified at the tender stage. The attempt was unsuccessful, however, because of adverse ground conditions. Hochtief abandoned the exercise when the pilot hole had reached a point 28m above the tunnel crown.

 

8. SSE Operates the Scheme

Takeover

[79]      SSE issued the takeover certificate to Hochtief in December 2008. In his accompanying letter, Sandilands stated:

“I confirm that I am satisfied that the power station and associated tunnels is now sufficiently complete for safe operation and is taken over by the Employer on 18 December 2008. The Employer also takes over the headrace and tailrace tunnels. The remainder of the works is taken over when sufficiently complete.”

 

[80]      The defects date was two years later and therefore became 18 December 2010. The parties expected, however, to carry out the final inspection in summer 2010 to avoid any problems with winter weather.

[81]      SSE took over the remaining works on 5 February 2009. Hochtief received a £2 million bonus for achieving early completion. SSE also paid Hochtief £5 million to close out all outstanding variations.

 

O & M Manuals

[82]      Hochtief sent 52 lever arch files and a CD Rom to SSE in autumn 2008. They contained the draft operation and maintenance (‘O&M’) information for the scheme. As they were too detailed for everyday use, SSE intended to prepare concise operational notes for its engineers.

[83]      Neither the manuals nor the notes had been completed at the time of the collapse. The manuals did not instruct SSE to check the net head. Claire McConnell, SSE’s Operations Systems Manager, intended to include a section on needle opening within the notes.

 

National Grid Tests

[84]      In order to export electricity to the National Grid, the Glendoe scheme had to pass various tests. These were undertaken by Andritz using its own specialist instruments. It completed the efficiency tests in January 2009. The governor and excitation systems required additional testing. The excitation system was not fully compliant even by the time of the collapse.

 

Monitoring

[85]      Glendoe is designed to be an unmanned operation, controlled from the GOC. An engineer sets the load and presses the start button. That activates the turbine governor, which in turn opens the needle valves. The scheme normally achieves its maximum output of 100MW at 80% valve opening. It depends in part on the level of water in the reservoir.

[86]      Electronic sensors monitor every aspect of operation. The scheme will automatically shut down if they detect a critical fault. The sensors also send signals to the GOC. To send all of them, a total of 1,700, would overwhelm the engineers with information. After detailed discussions the parties agreed to relay 179 signals, of which 140 yielded an audible alarm. The list did not include signals in respect of the needle valves or head loss.

[87]      When SSE engineers visited Glendoe to carry out maintenance or to investigate any problem, they could view the touchscreen display on the turbine governor. It showed the needle opening position, the power output, the net head and the flow. It could also show the trends in the performance of the plant.

 

Glendoe’s Role within the SSE Fleet

[88]      SSE intended Glendoe to generate electricity quickly to cover peak periods of demand. It expected the scheme to run for 20% of the year, depending on the level of rainwater in the reservoir. It only operated on a few days in June 2009, for example, because the reservoir level was very low. The scheme generally performed in line with expectations until SSE shut down the scheme. On 23 June it informed Hochtief that the scheme had achieved its availability guarantee of 99% during the first six month period.

 

Odd Readings

[89]      Sometime in late May or early June 2009, Sandilands was in the control room at Glendoe when he noticed a net head reading of 584m. As that was very low he mentioned it to his colleague Brian Still, who was with him at the time. Still said that it was probably a calibration error. Sandilands accepted the explanation because: “failure was not in my mind at all, this was a brand new hydro scheme”. In cross examination both men conceded that, with the benefit of hindsight, it had been a mistake not to investigate the cause of the reading.

[90]      There were other odd events in this period. Still attributed them to a ‘glitch’ in the control system. On 1 June he asked Andritz to investigate swings in output from the plant. In response to a query, on 3 June Still confirmed to Andritz that no alarms had sounded. Andritz did not provide any further response. This may have been because on this second email Still had ticked a box marked ‘no reply required’.

[91]      HM the Queen performed the official opening ceremony at Glendoe on 29 June. Initially the turbine failed to start that morning, but it did so on the next attempt. The following day Jacobs heard unusual thumping noises over a period of two hours. Still asked Andritz for an explanation on 2 July, while Jacobs put the same query to Hochtief on 7 July. Neither company responded.

[92]      On 8 July Still made a note about the odd output readings, which he continued to attribute to either a calibration or an instrumentation error. There was an unusual net head reading on 10 July. Andritz told Still on 24 July that it was unlikely to be a software error. It suggested checking the pressure indicator. Still had not done so before the collapse. During cross examination he agreed that he regretted not acting more promptly: “Had I made the check earlier and advised others earlier, it would have helped the situation.”

[93]      On 30 July, the turbine took five minutes to reach 100MW. One expert, Mr McWilliams, thought that a collapse may have occurred while the machine was shut down, resulting in a complete blockage, but the water burst through after five minutes. The machine operated for three hours that day and almost reached maximum output.

[94]      On 2 and 4 August the scheme failed to achieve its load despite the needle valves being fully opened. The judges of the Saltire Society engineering awards visited the scheme on 4 August. During their visit Sandilands of SSE and Wallace of Hochtief noticed a ‘sediment plume’ of discoloured water being discharged from the TRT into Loch Ness. Neither man took further steps to investigate.

[95]      When Wallace was asked why he did not discuss it with Sandilands, his first response was “no comment”. He then added that it was difficult to discuss matters openly in the presence of the Saltire judges. He also stated that these were issues for Andritz.

 

9. The Collapse & Its Aftermath

Detection

[96]      On 5 August a GOC engineer started the Glendoe scheme and set a load of 100MW. After 25 minutes he noticed it was only operating at 80MW. He monitored it for another 40 minutes. As it did not achieve the set load he shut down the turbine.

[97]      Mr MacKnocher is the engineer in charge of the SSE fleet in that area. He was at Foyers power station with a colleague when he received the news from the GOC. It did not occur to either man that there had been a tunnel collapse. Both suspected that the cause of the problem lay with the turbine governor. MacKnocher went to Glendoe to investigate. He asked the GOC to start the turbine and monitored the test run. Having looked at the data on the control room screen he made the following note:

“Glendoe reduced MW. Low headrace pressure. Steadily increased loss since March/April.”

 

Initial Strategy

[98]      Senior SSE personnel met the following day in Perth. They considered various issues, including remedial works, liability and insurance. Although the cause and scale of the problem were not known, Sandilands described the mood as “shell-shocked”.

[99]      At the close of the meeting SSE decided: (i) to review all operating data, (ii) to send down divers to inspect the reservoir intake screens, (iii) to dewater the tunnel, (iv) to send in inspection parties, and (v) to set up a steering group for the recovery project.

[100]    Subsequently SSE instructed AMCO to make the site safe and to carry out all necessary work to facilitate the inspections. It also engaged Donaldson to assist in the investigations. SSE made both appointments under existing framework agreements that it had with these companies.

[101]    Trevor Williams made a handwritten note of the meeting on 6 August. It contains the entry: “Give direct to Hochtief now? – Neil says no”. Sandilands could not recollect what he meant by that comment. When Hochtief saw that entry at a much later date it placed a sinister interpretation upon it. It inferred that SSE sought to prevent it from carrying out the remedial works from the outset. I conclude that it does not bear that inference. It simply reflects a practical approach. In order to preserve the insurance position, investigations had to take place promptly. It was convenient for SSE to take charge. Most Hochtief staff had left the site. It might have taken weeks to transfer responsibility for site safety back to Hochtief.

 

SSE Informs Hochtief

[102]    Julius Wallace and Jonathan Bird were the first Hochtief employees to learn of the problem. Sandilands telephoned them while they were lunching at the Lovat Arms in Beauly on 6 August. More formal communications followed later that day.

[103]    First, the contract supervisor, Appleby, issued Defect Notice D053 to Hochtief. It stated (a) that there had been an increase in head loss in the HRT, (b) that the plant could not run at its rated output, (c) that SSE had shut down the scheme to prevent further damage, and (d) that the support along part of the tunnel length “would appear to be inadequate”.

[104]    Second, Sandilands wrote to Hochtief acknowledging that it would find it difficult to meet the tight timescale laid down in the contract to rectify a defect resulting in a shutdown. He indicated that he was mobilising AMCO to make the area safe and to facilitate correction of the defect as soon as possible. Hochtief responded four days later. It stated that it did not accept liability at that stage and that it was preparing to mobilise.

 

Tunnel Inspections

[105]    The divers who inspected the intake screens found no evidence of any blockage. Dewatering began on the evening of 6 August. It took six days to empty the tunnel and reservoir.

[106]    The first inspection party entered the HRT on 14 August. It comprised Andy Gregory and Bill Bryce of SSE, Paul Owen from AMCO and Professor Sloan of Donaldson. Sloan heard and felt rumbles from deep inside the mountain as he stood at the top of the access shaft at the reservoir end. The sound of blockfall continued as the party walked down the tunnel.

[107]    They came across two minor collapses before a substantial pile of debris appeared out of the gloom. This was the upper end of the main collapse. It was 2km from the reservoir end of the tunnel. The debris pile almost totally constricted the tunnel. Professor Sloan noticed a gash around the tunnel perimeter and a void above, from where he believed the debris material had come. He could not ascertain its extent. Bryce prepared the report of the inspection. It concluded that the tunnel was unserviceable.

[108]    A week later representatives from SSE, Hochtief, Jacobs and the insurers conducted a second inspection of the upper part of the HRT.

[109]    The first inspection of the section of the tunnel below the debris pile took place on 26 August. It was a much more risky enterprise. Any release of water or debris down the steep incline could have resulted in serious injury or damage. Appropriate safety measures were therefore put in place. As the inspection party walked up the tunnel, they encountered deep, wet sand in the invert. Together with the blocked section, the total length of the deposits extended for 600m, much further than expected.

[110]    Afterwards Bryce made a ‘Heath Robinson’ sketch on a whiteboard showing the blocked section to be 270m long. The sketch was photographed and annexed to the inspection report. For several months it became the accepted measurement. The question of liability was again raised and disputed at the discussion following the inspection on 26 August. Hochtief maintained that it was an employer’s risk event. SSE asserted that it was a contractor’s risk event. They have maintained these positions to this day.

[111]    Another inspection took place on 16 & 17 September. Afterwards, the representatives of SSE, Hochtief and Pöyry discussed the technical solution. The consensus was to build a by-pass tunnel with a reinforced lining and a concrete invert. It would be positioned 50m away from the existing tunnel. In addition a downstream access tunnel would allow concurrent works to take place above and below the blocked zone.

[112]    Dr Wilhelm participated in this inspection and produced a report dated 19 September. He had by then left Hochtief and joined Lahmeyer. Apart from the collapse itself, he thought that the tunnel was “perfectly ok” requiring perhaps minor repairs. He regarded the rock mass as competent and strong. Some material had been washed out of shear zones during operation, but in his view that did not give rise to any stability problems past or future.

 

10. Relations Cool

Attempts to Reach a Consensus

[113]    Initially the parties’ communications were amicable in tone. For example Sandilands wrote to his counterpart Hans Dieter Imhof: “If a major rockfall has occurred it looks as if it will present a challenging job for whoever pays for it!” Gradually tensions developed.

[114]    The early exchanges sowed the seeds of much that followed. On 31 August Appleby instructed Hochtief to proceed with remedial works in terms of Clause 43.1. Hochtief replied that it could not commence works without the approval of its insurers. SSE wished to begin the recovery project as soon as possible. Hochtief insisted that detailed investigations should take place first.

[115]    On 7 September Donaldson produced a report estimating that the length of the collapse was nearer 100m than 270m. SSE did not relay that information to Hochtief until the following year. When it learned of the matter, Hochtief’s personnel were aggrieved. They believed that this was critical information which might have influenced the choice of the repair option.

[116]    The parties held the first meeting of consequence in Perth on 11 September 2009. Hoekstra was the senior officer representing Hochtief. He was receptive to SSE’s proposal to share the costs on a 50:50 basis, however, he did not have the authority to bind Hochtief. Three days later his colleagues decided that SSE should pay for the repairs. They were not prepared to contribute unless and until it established liability against Hochtief.

[117]    The parties had another ‘without prejudice’ discussion about cost sharing after the inspection on 17 and 18 September. SSE sent an email to Hochtief on 18 September stating that Hoekstra had agreed to share costs at the Perth meeting. He thought this was an attempt to ‘bounce’ Hochtief into a deal.

 

Conference Call on 25 September

[118]    Hochtief put Dr Frank Hormes in charge of its response to Glendoe. His first involvement with SSE occurred in a telephone conference call on 25 September. He proposed that both parties and their insurers should nominate an expert panel of engineers. No works would begin until they had completed their investigations and analysed the results. SSE had misgivings about the proposal. The contract did not allow for such a panel. More importantly SSE feared that it would lead to a stalemate. Hoekstra accepted that was a legitimate concern.

[119]    On the same day Hochtief set out the three pillars of its position in writing to SSE. First, the collapse was an employer’s risk event because SSE had issued the completion certificate following inspection. Second, Hochtief had started to remobilize to repair the collapse. Third, it wished further investigations to be undertaken to the satisfaction of its insurers.

 

Meeting at Edinburgh Airport on 5 October

[120]    SSE, Hochtief and Jacobs met at Edinburgh Airport on 5 October, along with the loss adjusters and their professional advisors. The discussion ranged over many technical, practical and commercial issues. The parties could not agree a way forward. The familiar impasse remained. Hochtief wished further investigations, while SSE wished to commence the remedial works.

[121]    Following the meeting, SSE instructed (a) Jacobs to investigate the mechanism and extent of the collapse; (b) Fugro to carry out geophysical investigations; and (c) the BGS to undertake further mapping, particularly in the CFZ.

 

SSE Meets HSE on 12 October

[122]    SSE met Dr Lamont of the Health and Safety Executive (‘HSE’) on 12 October. He was HSE’s expert on tunnel works. SSE wanted to find out if it would impose any conditions on the recovery project. Lamont singled out the invert and the void as two areas of concern. He wished a concrete invert along the entire HRT. He also wished the void to be stabilised, otherwise it might ‘propagate’ to the surface and create a danger to gamekeepers and hill walkers.

[123]    As a result of a misunderstanding Hochtief was not present at the HSE meeting. SSE believed that Hochtief thought it premature to discuss the repair works with HSE. In fact Hochtief had been willing to discuss the investigation works. This incident influenced the attitude of both parties. It reinforced SSE’s view that Hochtief was not committed to finding an expeditious solution. It underscored Hochtief’s view that it was being excluded from the recovery project.

 

Two Further Meetings on 20 and 21 October

[124]    SSE and Hochtief met in Glasgow on 20 October and again at Edinburgh Airport the following day. SSE expressed its frustration at the lack of progress. Hochtief agreed to price certain parts of the recovery project, subject to the completion of investigations and the production of a final design. For the first time SSE voiced the possibility of instructing an alternative contractor to do the work. On 27 October SSE issued three defects notices to Hochtief: D054 related to the aqueduct tunnel, D055 to the HRT, and D056 to the TRT.

 

Essen Design Workshop on 29 October

[125]    Sandilands and Brand of SSE attended a design workshop in Essen on 29 October at which Hochtief presented seven remedial options. The options included such elements as a bypass tunnel, a downstream access tunnel and a concrete invert. Depending on the option chosen, the programmes varied in length from 22 to 31 months.

[126]    Hormes believed that SSE was happy with the proposals. After the meeting he wrote to SSE:

“As agreed … design works can only start once the results of the investigative works and their interpretation are available and their impact on the design parameters have been determined by the experts under guidance of SSE.”

 

[127]    Sandilands and Brand were sceptical about Hochtief’s commitment to the recovery project. As they left the meeting, they recall Hormes making a remark along the lines of “no‑one in Hochtief is allowed to spend any money on this.” Sandilands described the workshop as ‘critical’ to his thinking. It crystallised his doubts about whether Hochtief intended to carry out the remedial works. Brand held a similar view: “there was an awful lot of talk going on and not a lot of meat coming out of it”.

[128]    Sandilands and Brand reported their concerns to a meeting of the SSE steering group held on 2 November. Ian Funnell, SSE’s Director of Major Projects, regarded the documents they had brought from Essen as “very light”. He had met Hoekstra on several occasions and was likewise concerned that there had been no substantive progress. Funnell believed that Hochtief had not committed staff or resources to the site. Against that background, SSE decided to approach four other contractors to undertake the remedial works. It still preferred Hochtief to carry out the works, however, because of its knowledge and experience of the site.

 

Hochtief’s Proposals

[129]    During this period Hochtief floated various proposals for discussion. It suggested sharing the investigation costs on a 50:50 basis if the joint insurers paid its share. It reiterated its willingness to mobilise to site if SSE paid it to do so.

[130]    Most importantly, on 12 November it made a formal offer to construct a downstream access tunnel (‘DAT’) in return for payment on a cost reimbursable basis, plus a 12.5% fee. The offer was open for acceptance until 29 November. SSE did not accept the offer. It wished to reach agreement on a solution that encompassed all the remedial works.

 

SSE Issues an Ultimatum

[131]    On 7 December SSE wrote an ‘ultimatum letter’. It required Hochtief to provide a programme for the remedial works within 14 days and to agree a 50:50 costs sharing until liability had been determined. Hochtief replied the following week. It refused to comply with the conditions, denied that it had failed to make progress and suggested a further meeting before Christmas. SSE did not see the point of another meeting. From its perspective, Hochtief had failed to grasp its last opportunity.

[132]    On 23 December SSE invoked the Availability Guarantee and ordered Hochtief to pay £1M in damages. On 18 January 2010 it notified Hochtief that it intended to appoint BAM Nuttall to undertake the recovery project.

 

11. Specialist Investigations

Jacobs

[133]    In early October 2009 Jacobs produced an outline of the investigations it thought were necessary for the remedial works. It proposed that there should be surface mapping, aerial photography, seismic topography and the drilling of boreholes at specified locations. Hochtief, Donaldson and the insurers revised the draft document to add the investigations that they sought.

 

British Geological Survey

[134]    The BGS is a research establishment funded equally by the public purse and the private sector. During the course of the TBM drive, two BGS geologists visited the tunnel several times. They described themselves as ‘geo-tourists’ and were enthralled with what they saw: “it was like looking at a textbook”.

[135]    After the collapse, SSE instructed BGS to prepare a fresh geological map of the CFZ. Dr Smith led the team that carried out the task, which took over four years to complete. The BGS concluded: (a) that the CFZ is a complex fault structure with multiple zones of fractured rock over a short distance; (b) that it intersects the HRT at the collapse zone; and (c) that the nature of the fault rocks in the collapse area is unknown. The BGS team produced a 3D model showing its findings.

 

Donaldson

[136]    Professor Sloan was in overall charge of the Donaldson team that mapped the HRT and the TRT from autumn 2009 until August 2011. It found both tunnels to be in sound rock, but compiled a list of locations where it considered that the support did not comply with the Works Information. Professor Sloan made the final decision about which of these ‘secondary tunnel features’ should be classified as defects.

 

12. The Recovery Project

Commercial Pressures on SSE

[137]    SSE was losing revenue of £50,000 per day while Glendoe was out of service. It therefore had a powerful incentive to bring the scheme back into service as soon as possible. There were other commercial factors that pointed in the same direction. SSE’s insurance cover for business interruption expired on 5 February 2011. Any claim it had against Hochtief for lost generation income liability was capped at £1 million.

 

SSE Selects BAM

[138]    The Royal BAM group of companies had some knowledge of the scheme. It had made an unsuccessful bid for the original works. Its divisional director, Mr Henderson, had advised SSE on investigations after the collapse.

[139]    In late November SSE invited BAM to tender for the recovery project. For a contract of this size and complexity, BAM would normally have taken 20 weeks to prepare a bid. In this instance, it only had two weeks. In its estimate of 7 December: (i) it proposed a 830m long bypass tunnel; (ii) it set a completion date of late 2010; (iii) it gave an indicative price of £30 million; and (iv) it stipulated that the contract should be based on Option E of the NEC conditions.

[140]    SSE formally appointed a BAM joint venture to undertake the recovery project in March 2010. The three companies within the joint venture were BAM Ritchies, BAM Nuttall and Wayss & Freytag. Together they had expertise in geo-technics, civil engineering and tunnelling.

[141]    SSE did not rely solely on BAM’s expertise.  Throughout the recovery project SSE acted on the advice of Jacobs as the designers, Donaldson as the technical peer reviewer and independent design checker, Gardiner & Theobald as the project cost consultant and Ash Consulting Service as the environmental consultant.

 

Choice of Option E

[142]    The NEC suite of contractual options is designed to meet different situations. SSE had entered into a fixed price contract with Hochtief on the basis of Option A. With regard to the recovery project, SSE entered into an Option E contract with BAM. That is also known as a ‘cost reimbursable’ contract. It entitled BAM to recover all its costs, together with a percentage fee to cover overheads and profit. I shall say more about this model form of contract below.

 

BAM Mobilises to Site

[143]    BAM established itself on site in early 2010. It drilled boreholes to investigate the crown and the void. On 26 May, it used an electronic theodolite to determine with accuracy the length of the blocked section of tunnel. They discovered that it was 71m long. It was some time before it conveyed this information to Hochtief, who regarded this as a “fundamental piece of information with very significant implications” for the repair works: email to SSE dated 10 July 2010.

 

Scope of the Remedial Works

[144]    The main elements of the remedial works were: (a) a 605m bypass tunnel around the collapse; (b) a 550m downstream access tunnel; (c) a concrete invert along the entire length of the HRT; (d) secondary lining repairs in both the HRT and the TRT; (e) repairs to the dam outlet; and (f) re‑commissioning of the plant. The design constantly evolved throughout the course of the contract and was not finalised until September 2011. That is reflected in the fact that BAM put forward 579 compensation events (changes to the works information).

 

Contract Changes

[145]    In January and again in November 2011 SSE and BAM signed supplemental agreements. On the first occasion they varied the contract to close off disputed costs. On the second occasion they shifted from a cost reimbursable to a £109 million fixed price contract. SSE had pressed for this change to limit the escalation in costs. BAM was only prepared to do so after the finalisation of the design in September.

[146]    There was a noticeable improvement in the mood on site after the change from Option E to Option A. Mr Brand of SSE felt that an invisible weight had been lifted from his shoulders. BAM was able to progress the works without constant challenges from SSE. Subsequently, the parties agreed various compensation events totalling £5 million, which resulted in BAM’s final account of £114 million.

 

13. Why did the Tunnel Collapse?

The Short Answer

[147]    The HRT collapsed because there was not enough support: poor rock conditions coincided with insufficient shotcrete and rockbolts. That is the short answer to the question of why the tunnel collapsed.

 

A Longer Answer

[148]    There are formidable difficulties in providing a longer answer, because no full investigation into the cause of the collapse took place. It ceased to be a priority issue once SSE elected to construct a bypass tunnel.

[149]    It is convenient to summarise the current state of knowledge about the collapse. Two facts are known: (a) the blocked zone is 71m long, and (b) the collapse developed progressively from at least 12 April 2009. Two other facts remain, however, unknown: (a) the length of the crown that collapsed, and (b) the dimensions of the void.

[150]    Expert views on the length of the crown that collapsed diverge widely. Professor Sloan suggests a 71m collapse, while in Dr Büchi’s opinion, it is only 8m - 10m. I agree with Professor Grøv that it is “pure guesswork”. In my view it is unlikely that the collapse occurred over the whole length of the blocked zone for two reasons. First, it would be unprecedented in the history of tunnel collapses. Second, the de-watering curve would have been materially different to reflect the greater difficulty presented by such a long blockage. Equally the large volume of material in the debris pile suggests a collapse longer than 10m. As to the dimensions of the void, the estimates range from 2,374m³ to 13,000m³.

[151]    Having regard to the weight of the expert evidence, I incline to the view that it is impossible to determine the cause for the collapse. As Büchi put it, the collapse was “a geological accident”. Smith is not an engineering geologist. He therefore cannot offer a view about the cause of the collapse. But from a purely geological perspective, he said: “If you wanted to build a tunnel in this area, you couldn't have picked a worse spot in terms of the structural complexity”.

[152]    I conclude that the most likely explanation, which is neutral on the question of fault, is as follows. (1) The CFZ consists of interconnected faults of thin single shears with good rock in between. (2) The weak rock deteriorated and lost its strength when submerged, a process Professor Stille referred to as “slaking”. (3) The flowing water washed out areas of erodible rock. (4) The erosion progressed and opened up larger seams. (5) The eroded material was progressively deposited as sediment over a significant length of the HRT. (6) The HRT lost stability and the tunnel collapsed. (7) Dewatering caused further erosion.

 

14. Witnesses & the Burden of Proof

[153]    Before turning to consider the issues, I wish to comment on the quality of the evidence. The lay witnesses were of a high calibre and there were very few issues of credibility and reliability. The experts were impressive and in many instances were eminent in their field. I found it disappointing, if not surprising, that there was so little accord among them. Almost every issue in the case was disputed. The tunnelling experts did, however, agree that it was appropriate (a) to excavate the tunnel by means of a TBM, (b) to adopt an unlined design concept, and (c) to use the observational method of rock classification.

[154]    Senior counsel for SSE invited me to scrutinise the evidence of Professor Broch with care, because of his dual involvement as an expert and as an adviser to Hochtief. Senior counsel for Hochtief issued the same invitation in respect of Professor Sloan. I formed the view that both witnesses did their level best to assist the court and did not give biased evidence.

[155]    Both sides made various objections in the course of the proof. On some occasions, Hochtief lodged a written note at the outset of the evidence of individual SSE witnesses. Most objections asserted that the witness was not properly qualified to give expert testimony, either because he was truly a factual witness, or because he was not truly independent. In almost all cases I allowed the evidence subject to competency and relevancy. The objections did not figure largely in the final submissions.

[156]    There were several sessions in the proof where I heard concurrent evidence from the experts. This procedure, known colloquially as ‘hot-tubbing’, involved several experts being present in court at the same time. I chaired a discussion between them with a view to crystallising their respective positions. I found it a valuable way of focussing on the main issues and assessing the quality of their contributions.

[157]    With regard to the onus of proof, SSE relies on the fact that a number of key individuals did not give evidence, including Taylor (Hochtief); Kellaway, Williams and Appleby (all Jacobs); and Agar (Pöyry). SSE submits that Hochtief cannot discharge the burden of demonstrating that it exercised reasonable skill and care in the decision‑making process.

[158]    That is an unduly narrow approach. I do not know and shall not speculate on the reason why Taylor was not called. The evidence as a whole, however, was comprehensive and comprehensible. I am satisfied that it provides me with the basis to answer the key questions. Questions of onus are seldom decisive once evidence has been heard: Salt International v Scottish Ministers [2015] CSIH 85.

 

15. Was there a Defect at Takeover?

Scope of the Dispute

[159]    Under Clause   80.1, at take over SSE assumed the risk for employer’s risk events, which included:

“Loss of or damage to the parts of the works taken over by the Employer, except loss or damage occurring before the issue of the Defects Certificate which is due to… a Defect which existed at take over …”.

 
[160]    Clause  11.2 (15) defines a defect as:

“a part of the works which is not in accordance with the Works Information or a part of the works designed by the Contractor which is not in accordance with the applicable law or the Contractor’s design which has been accepted by the Project Manager.”

 

[161]    SSE contends that both limbs of the definition are satisfied. The scheme could not “provide reliable service without requirement for major refurbishment or significant capital expenditure”, contrary to Clause 6.3.2 of the Works Information. The scheme did not conform with the accepted design, because Hochtief failed to install the requisite level of support to prevent the erosion of erodible rock during operation.
[162]    Hochtief resists these arguments. It relies on Option M, which states:

“The Contractor is not liable for Defects in the works due to his design so far as he proves that he used reasonable skill and care to ensure that it complied with the Works information”.

 

Was The Scheme Contrary to the Works Information?

[163]    I reject SSE’s first submission, because in my view Option M placed an important brake on liability. Hochtief did not guarantee the works. Instead it accepted the familiar and lesser obligation of “reasonable skill and care”. No doubt that had practical consequences. Hochtief’s assumption of a more limited degree of risk would have been reflected in the contract price and the level of its insurance premiums.

[164]    If SSE’s interpretation was correct, it would mean that Hochtief had an overarching obligation to provide tunnels suitable for their purpose. That would impose strict liability on Hochtief. It would effectively rob Option M of its meaning. It would also make redundant some of the other provisions regarding repair and maintenance.

 

Was The Scheme Contrary to the Design?

[165]    SSE maintains that Hochtief failed to fulfil its obligation to shotcrete all erodible rock in terms of table 11 and chapter 6.1.5 of the tunnel support methodology. It should have installed class III or IV support wherever it found erodible rock. Professor Sloan explained the rationale as follows: 
 

“That’s why support is not selected at the face, there is a system in place that protects the project, the design is taken into the tunnel to be implemented.”

[166]    In my view there are formidable difficulties with that approach.

a)   These two isolated provisions run counter to the overwhelming weight of the other contractual provisions, which called for the exercise of engineering judgement in the tunnel. I mention these in para 170 below.

b)   The term “erodible rock” is itself vague and required interpretation at the face. Professor Sloan himself would not have protected narrow seams that did not pose a threat to stability.

c)   If SSE is correct, determination of the “overall excavation class” on the REC sheet should have been a mechanistic process. It depended on the worst feature ticked.

d)   SSE would have had to pay for a far greater level of support than that which was actually installed.

e)   It would no longer be possible to have an unlined tunnel in similar geology. The classification would inevitably result in a lined tunnel.

f)    The heading of chapter 6.1.5 is ‘Water Outflow and Loss’. It addressed the different issue of high permeability zones that might endanger individuals and plant. Because the rock conditions in the tunnel were so dry, this provision was not engaged.

[167]    Having regard to these factors, I conclude that there was no mandatory requirement to shotcrete all erodible rock. The engineering geologists had to exercise judgement in considering the integrity of the tunnel. It was a holistic exercise. They had to assess the presence of any erodible material, together with the dip direction, orientation and width of any faults.

 

16. Reasonable Skill and Care

[168]    SSE submits that Hochtief failed to exercise reasonable skill and care. As a result it incorrectly classified the rock and installed inadequate support for the long‑term operation of the tunnel. SSE also submits that Option M only provides a limited defence. It applies to design defects, but not to implementation defects.

[169]    I shall split my discussion of this issue into three sections. First, I shall consider what the parties agreed about the classification process. Then I shall consider whether Hochtief implemented that agreement. Finally I shall look at the joint inspections that took place in 2008 prior to watering up.

 

What did the parties agree about classification?
[170]    The contract documents have a consistent theme. Classification was to be jointly carried out at the face:

·    Reference ground conditions report: “During construction, the support requirements will be decided at the face, based on the actual conditions”.

·    Works Information: (i) “Following each excavation cycle, the contractor maps the face, crown and sidewalls to enable the classification of the ground in accordance with the rockmass classification system”. (ii) “The contractor agrees the rockmass parameters with the project manager and the support class is agreed prior to its installation.”

·    Design drawing D201: (i) “Rock supports to be installed to the extent required to meet the rock conditions encountered”. (ii) “Rock support and surface treatment may be modified as considered necessary to adapt to the actual geological site conditions at the work site.”

·    The baseline ground conditions report and the design statement for the tunnel are to similar effect.

[171]    I conclude that the parties clearly agreed the approach to rock classification and support. The engineering geologists should jointly determine both matters within the tunnel. This collaborative approach had great utility. It took advantage of the experience of all involved. Further, as Sandilands recognised:

“it was beneficial for all parties for the classification of the ground to be checked as it went along because the consequences of defects being raised at a later date would create difficulties for both parties”.

 

It also ensured that SSE only paid for necessary support.

 

Standard of Mapping

[172]    SSE submits that Hochtief systematically misclassified the rock. It relies largely upon Professor Sloan’s opinion that the engineering geologists missed defects at 114 locations (96 in the HRT and 18 in the TRT). As Hochtief failed to identify defects at these locations, he infers that it failed to identify major ones in the blocked zone. He also says that the rock conditions in the bypass tunnel provide an analogue that enables him to say that similar ones prevailed in the HRT and were missed by Hochtief.

[173]    The other SSE experts were less robust.  Professor Grøv, for example, paused for a significant period before saying that the mapping did have indicators of a potential collapse. In the adjudication he had given a different view. He said that the Hochtief mapping sheets “do not explicitly record weak zones of any significance. Thus from the mapped geology alone, it is difficult to identify why a major collapse occurred.” (report paragraph 3.11)

[174]    In deciding whether Hochtief exercised reasonable skill and care when carrying out the tunnel mapping, I have regard to several factors. First, mapping a TBM tunnel is a difficult exercise. Because of its smooth surface “details may get lost” (Professor Grøv). This is compounded at Glendoe, as it contains: “some of the hardest rocks to classify under ideal circumstances” (Professor Sloan, email 20 December 2010).

[175]    Second, the Hochtief experts variously described the standard of the mapping as being of “good quality” (Dr Palmström), “top quality” (Dr Büchi), and that it “complied with international standards” (Professor Broch). Accordingly, a responsible body of opinion within the discipline of tunnelling supports Hochtief.

[176]    Third, I accepted the evidence from all members of the TBM crew that Taylor was an experienced engineering geologist, who diligently performed his role. He took extra care at the predicted location of the CFZ, but saw nothing. He did not have the detailed information later assembled by BGS. The fact that Taylor noted weak rock, such as mica schist and kakerite, on some REC sheets reflects the care he took. I am satisfied that in his judgement they did not merit further protection. Why go to the trouble of identifying such features and then decide to do nothing about them?

[177]    Fourth, Jacobs and Pöyry agreed with Taylor’s classification. They had the opportunity to query his decisions (a) in the tunnel, (b) when the REC sheets were put into the geological database, and (c) when they signed the REC sheets. They chose not to do so.

[178]    Dr Büchi, Professor Broch and Dr Smith all thought that the rock conditions of the bypass tunnel are not relevant. Dr Smith said that while the gross geological features in the two tunnels will be similar, there may well be variations in the orientation of faults and the way in which they interlink. In any event, the information was not available to Hochtief.

[179]    I disagree with Professor Sloan’s description of joint mapping as meaning that two geologists go into a tunnel together, carry out the mapping together, fill in the appropriate sheet and sign it off on the day as a joint map. Prior to the collapse he provide a report to Hochtief in which he referred to the parties following a joint approach (September 2007).

 

Joint Inspections in 2008

[180]    Many experienced tunnellers scrutinised the HRT both during and after the TBM drive. They included engineering geologists, tunnel designers, engineers and the TBM crew. They were actively looking for problems. None of them saw signs of faults that might threaten tunnel stability. None recommended the installation of a higher level of support at particular locations.

[181]    Hochtief, Pöyry, Jacobs and SSE inspected the whole tunnel on a metre by metre basis in 2008, but only uncovered several minor issues which were resolved prior to watering up. I reject SSE’s contention that “the fatal error occurred in October 2008”, because no engineering geologists were present at the pre-watering up inspections. To do so focusses on one inspection out of many. Mr Fawcett, a distinguished tunneller, wrote in his first report: “There is no recorded evidence of passing through any feature that would potentially cause the catastrophic collapse that has occurred.”

 

Three Miscellaneous Issues

[182]    It is convenient to deal here with three other issues related to reasonable skill and care.

 

(a) Did the REC Sheets Address Long Term Hazards?

[183]    SSE contends that the REC sheet only addressed the hazards during excavation. That is a startling proposition. Why would Hochtief construct a tunnel with no thought to its long term stability? SSE relies on an internal Hochtief email that Taylor sent to Hormes on 8 October 2010. I shall set it out in full:

“The extent of the collapse suggests that a significant high hazard zone was encountered in the tunnel. Such zones would normally require a high degree of installed support involving full periphery lining and steel ribs consistent at least with the excavation Class  IV defined by the Designer.

However, the classification system imposed by the Designer for the selection of the support to be installed was based on assessing the rock for evidence of geotechnical indicators and properties that posed specified hazards for the excavation of the tunnel. Having identified these geotechnical properties, the excavation class was specified which then corresponded to a particular support design.

During the excavation of the HRT this system was rigorously applied and subsequently verified by all parties. Furthermore, it is understood that the Owner’s Engineer undertook verification mapping, applying a conventional rock mass classification system (Q-System) as a check on the adequacy of the classification or support being installed was insufficient for the conditions encountered.

In retrospect, the approved classification system does not adequately cater for a series of faults that link up in 3 dimensions … . In the HRT excavation, these boundary faults may appear as singular thin discontinuities that could only be categorised using the approved Excavation & Support Class  System, at worst as Class  II.

The fact that the independent support confirmation undertaken by the Owner’s Engineer appears not to have identified a highly unfavourable (Class  IV) zone gives confirmation that these were unforeseen ground conditions beyond the exposures created in the HRT.”

 

[184]    The reference to “specified hazards for the excavation of the tunnel” (emphasis added) does not in my view mean that the REC sheets failed to address long term risks. Taylor was simply explaining the process he had followed during the tunnel drive. He was analysing what had gone wrong with the benefit of hindsight. He was not conceding that the REC sheets solely addressed the risks at excavation

 

(b) Was Hochtief’s Classification System Deficient?

[185]    I hold that a different classification system would not have resulted in the installation of heavier support. Dr Palmström is eminent in this field and has himself devised a classification system. He stated that no system could have predicted the correct level support. That conclusion is confirmed by Jacobs’ use of the Q‑system. If it had not reached the same result as the Hochtief system, Jacobs would not have approved the REC sheets on behalf of SSE.

 

(c) Would Probe Drilling Have Made a Difference?

[186]    As the TBM approached the CFZ, the rock conditions remained very dry. There was therefore nothing to warrant the use of probe drilling, which is mainly used to detect water and gas. Several Hochtief employees who worked on the TBM (Kelly, Zimmerman, Keusch and Kaufmann) said that there were no concerns about tunnel safety, which is the usual reason for forward probing. In any event (a) probe drilling is not used to make decisions about permanent support; and (b) it would not have detected the CFZ, as the drill could not take cores. A number of witnesses pointed out that the TBM itself should have been a better detection device.  The cutters should have found it easier to penetrate the rock in poor conditions. The debris material on the conveyer would have included larger sizes of rock.

 

Conclusion on Reasonable Care

[187]    I am satisfied that Hochtief did exercise reasonable skill and care.  I reject SSE’s case as it depends on the accumulation and interpretation of all the data that has been obtained since the collapse. Put short, it is founded on hindsight.

 

17. Should Hochtief have Returned to Site?
[188]    Clause  82.1 states:

“Until the Defects Certificate has been issued and unless otherwise instructed by the Project Manager the Contractor promptly replaces loss of and repairs damage to the works, plant and materials.”

 

[189]    I construe that clause as creating a stand-alone regime. It detaches the immediate question of repair from the ultimate question of liability. Its purpose is plain. Wrangles about liability are postponed. If the damage is ultimately found to be an employer’s risk event, the contractor is entitled to payment. The repair works will be a compensation event: Clause 60.1(14). If, however, the damage is due to a contractor’s risk event then it must bear the costs: Clause 83.1.

[190]    From August to December 2009, Hochtief adopted an unvarying position. It was only prepared to carry out the remedial works in return for payment. It was not even prepared to share the costs on a 50:50 basis. I hold that by adopting this stance it breached its obligation under Clause 82.1.

[191]    As I hold that the collapse was an employer’s risk event, however, this had a minimal impact on the loss sustained by SSE. It still would have had to bear the cost of the recovery project.  The breach does have resonance in relation to the counterclaim under the principle of mutuality. If Hochtief was in breach of its obligation, it cannot sue SSE for breach of its obligation.

 

18. Contributory Negligence

[192]    Hochtief submits that SSE should have discovered the collapse by May 2009, when the extent of the damage would have been much less. Its experts, Vinnogg and Guttormsen, say that a reasonably competent operator would have monitored matters more carefully.

[193]    SSE’s experts, McWilliams and Honningsvåg, take a different line. They say that Glendoe is a complex hydroelectric scheme built to a bespoke design. Even by August 2009, it was not in a steady state. Testing was still taking place. There were therefore no normal operating conditions. It was understandable for SSE to attribute rogue observations to instrumentation errors. A major tunnel collapse within months of takeover was a highly unlikely event. Further, the signs were confusing. After each blip the scheme quickly returned to full output. Most importantly, Andritz was the specialist on site and was made aware of the odd readings. It did not advise SSE that there was an urgent problem.

[194]    I find the reasons of McWilliams and Honningsvåg to be cogent and convincing. I therefore hold that there was no contributory negligence on the part of SSE in failing to appreciate the significance of the odd readings.

 

19. Causation

[195]    SSE states that Hochtief should have installed at least class III support in areas of erodible rock. The experts agree, however, that only class IV support would have prevented the collapse. Professor Sloan said that he would err on the side of caution and install class IV. Even with hindsight “it is very difficult to say whether class III would have prevented the collapse”.  Professor Grøv, Professor Stille, Dr Palmström and Dr Büchi were to similar effect. I therefore conclude that even if SSE had established liability against Hochtief, it has not proved that class III support would have been enough to avoid the collapse.

 

20. The Costs of the Recovery Project

Introduction

[196]    Although I hold that Hochtief is not liable to SSE on the main grounds of action, I must determine what damages I would have awarded if the claim had been successful. I gratefully accept the parties’ invitation to deal with quantum on a headline basis.

[197]    There is one preliminary matter. I reject Hochtief’s contention that SSE failed to prove that SSE Generation Ltd was the corporate entity within the SSE group of companies that incurred the costs. There was unchallenged evidence to that effect.

 

The Position of the Parties

[198]    BAM expected to complete the remedial works in December 2010. In fact it did not do so until August 2012. No doubt that contributed to the dramatic increase in price. Its initial estimate of the contract price was £28 million. At that stage SSE authorised a total spend of £30 million. It approved a revised estimate of £85.7 million in January 2011. At the same time it agreed to pay BAM a bonus if it completed the works by 16 December 2011. On 28 October 2011 the Board reviewed matters and authorised £137 million to be spent on the project.

[199]    Hochtief submits that the price escalation cannot be justified, because the scope of the works did not significantly change, the recovery project was ‘over-engineered’ and BAM should have adopted a tunnel-through solution or constructed a shorter bypass tunnel. Hochtief points out that Dr Büchi characterised the result as “shocking”, while Dr Palmström said that it was “unbelievable”.

[200]    I shall anticipate my conclusion by stating that I hold that the cost of the recovery project was largely reasonable, subject to what I say about individual heads of claim in the next chapter.

 

General Principle
[201]    In Banco de Portugal v Waterlow & Sons Ltd 1932 AC 452, Lord MacMillan stated:

“It is often easy after an emergency has passed to criticise the steps which have been taken to meet it, but such criticism does not come well from those who themselves created the emergency. The law is satisfied if the party placed in a difficult situation by reason of the breach of a duty owed to him has acted reasonably in the adoption of remedial measures, and he will not be held disentitled to recover the cost of such measures merely because the party in breach can suggest that other measures less burdensome to him might have been taken.”

 

[202]    A number of factors align with that general principle and support the reasonableness of the cost of the recovery project. First, SSE had to act quickly to recover its asset. Second, it acted on the advice of its professional consultants. Third, BAM was only prepared to contract on the basis of Option E. Fourth, SSE instituted an exacting costs monitoring regime. Fifth, the remedial works did evolve over the course of the recovery project – Jacobs produced 75 reports generating over 500 compensation events. Sixth, no-one wished to put the scheme back into operation without being sure that it was fit for purpose. Seventh, it is not appropriate to compare the price of the original works with those of the recovery project.

[203]    Let me add a word on the last point. Mr Hindle said that the 1994 collapse of the Heathrow Express link is “burned on every tunneller’s mind in this country”. The recovery costs there were three times greater than the original contract price. In the present case the design, scale and constraints on the recovery projects were radically different from those faced by Hochtief.
[204]    I shall now consider various issues that arose in relation to damages.

 

Choice of Option E

[205]    Many employers view ‘cost reimbursable’ contracts with a degree of suspicion. They believe that they favour contractors, who are not under the same pressure to mitigate costs. But contractors also shy away from such contracts, because their accounts are subject to close scrutiny. They have to justify every penny spent. As Mr Leddie put it: “option E is an open book, not an open cheque.”

[206]    I am satisfied that Option E was the only appropriate model at the start of the recovery project. It allowed BAM to commence the works quickly and for the design to evolve. SSE would have taken many months to compile the works information necessary to conduct a tender process for a fixed or target price of contract. That would have caused significant delay. Alternatively, the contractor would have had to submit a very high price to reflect the risks of proceeding with an outline design and uncertainty about the ground conditions.

[207]    There are three further points. First, Mr Dalgleish of BAM said that it was only prepared to enter into an Option E contract. Second, Hochtief itself proposed to carry out the remedial works on that basis. Third, Hochtief led no evidence to establish that a different model form would have produced a lower price.

 

Costs Monitoring

[208]    SSE closely monitored the rising costs of the recovery project and sent regular reports to the underwriters. It held weekly meetings on site. It also held weekly meetings of senior staff at its headquarters in Perth, which was unprecedented. In addition, it took advice from external quarters.

[209]    Four former chairmen of the British Tunnelling Society – David Fawcett, David Court, David Donaldson, and Terry Mellors – provided advice to SSE. Their recommendations saved significant sums of money. For example they advised SSE (a) to install thinner shotcrete lining in the bypass tunnel, and (b) to remove part of the debris pile to reduce the length of the bypass tunnel. Court had worked on the Foyers Hydro Project in the 1970s. It involved tunnelling through a fault on the south side of Loch Ness. In a report dated February 2011 he concluded that the Glendoe works were being managed and resourced in an efficient manner.

[210]    SSE employed Gardiner & Theobald (‘G&T’) to act as cost consultants. G&T began its audit in February 2010. It constantly challenged BAM about efficient resource allocation. In 2011 G&T discovered that BAM had omitted to render an invoice for £1 million. As might be imagined, this caused a degree of consternation. SSE then engaged an accountancy firm, Henderson Loggie, to audit BAM’s cost management systems. Mr Leddie said that BAM telephoned him throughout the contract to complain about the stringency of G&T’s auditing.

[211]    Accordingly, I conclude that SSE continually tested matters from every angle. There was a robust project management process.

 

Expert Evidence on Costs

[212]    Hochtief relied mainly on the evidence of Hunter. He maintained that BAM used too many resources and too much labour. He estimated that the remedial works should have cost no more than £30.5 million. The SSE experts, Hackett and Goulding, regarded that as a woeful underestimate. In their view Hunter did not recognise the difficulties involved in recovering a failed tunnel.

[213]    Hackett and Goulding both have considerable experience of tunnelling. They approached the question of costs in a practical manner. Hackett made a detailed assessment of all the costs of the recovery project. He was a quietly impressive witness.

[214]    I found the evidence of Goulding to be equally persuasive. Instead of repricing the works, he obtained comparable rates from four different contractors to see whether BAM’s figures were reasonable in relation to certain key elements. He found that in some instances BAM’s rates were higher, in others they were lower. Overall he concluded that it did not make much difference. Goulding came to the view, however, that the sum paid to BAM could be reduced by £1.4 million to take account of possible inefficiencies in the design process.

[215]    I found Hunter’s evidence less cogent and compelling. In my view he failed fully to appreciate (a) the extent of uncertainty at the outset; (b) the scale and intricate sequencing of the work; and (c) the rigour of the costs monitoring regime. I agree with senior counsel for SSE: “If Mr Hunter is correct, SSE was ripped off by BAM”. I am satisfied that BAM focused on doing a good job and was not out to milk the contract for gain. In fact it only made a profit of 2% - 3%.

[216]    Two further aspects of Hunter’s evidence call for comment. First, in his reports he placed heavy reliance on Spon’s Engineering and Civil Pricing, which is an annual reference work for quantity surveyors. It contains indicative rates for labour, materials and plant hire. The section on tunnelling, however, recommends seeking the advice of a specialist contractor. Both Hackett and Goulding flatly rejected its use in relation to the complex recovery project at Glendoe. To my surprise Hunter agreed with them in court. He said that if he was pricing these works for a contractor, he would use his own firm’s information and experience rather than rely on Spon’s. In fact he discussed the pricing with Hochtief, because it had carried out the original project.

[217]    Second, Hunter referred to his involvement in an earlier tunnelling project at Fort William. Clearly that could have been a useful benchmark against which to assess the costs at Glendoe. After the first day of cross examination SSE elicited further information on the Fort William project overnight. Senior counsel for SSE then put to Hunter the next morning that the earlier contract was not a proper comparator. Hunter answered: “Yes, I wouldn’t disagree with that. It’s certainly not something upon which I would wish to significantly rely in saying to the court about experience I have in relation to such matters.”

[218]    Taken together, these points undermined my confidence in Hunter’s opinion. I do not accept his views (a) that BAM would have entered a fixed price contract of £30 million, (b) that the programme could have been completed in 72 weeks, and (c) that it would have cost (i) £7.5 million to tunnel through; and (ii) £9 million for a 153m bypass tunnel.

 

Health and Safety

[219]    It is worth emphasising that everyone involved in the recovery project was understandably concerned about health and safety. A major collapse had occurred. Its cause was unknown. There was a risk that another hazardous event could take place.

 

Betterment

[220]    As Brand accepted, the HRT is a better tunnel now than it was before the collapse. It has a smooth concrete lining. SSE has installed more accurate equipment to ascertain the trends for needle valve opening and the net head. It has also added an alarm for the head loss. As I am generally satisfied that the works were reasonable, I make no discount for these incidental benefits.

 

21. Individual Heads of Claim
[221]    In this chapter I shall comment on particular elements of the recovery project.

 

Bypass Tunnel

[222]    Hochtief argued that a bypass tunnel was not the correct solution, that in any event it was placed too far from the original tunnel and was too long. I reject each of these arguments.

[223]    Feasibility It is possible to tunnel through any blockage. In each case it is a matter of risk, cost and safety. SSE, BAM and Jacobs considered that a bypass option was the most sensible solution at Glendoe, given the degree of uncertainty about the size of the collapse, the extent of the void, and the stability of the rock. That remained their view in May 2010, when they found out that the collapse was only 70m long. HSE expressed several concerns about the risks of tunnelling through. I conclude that a bypass tunnel was a reasonable solution that minimised the risks.

[224]    Location BAM built the bypass tunnel 100m from the original tunnel. Various other distances were proposed: 50m (Hochtief at the Essen workshop), 40m (Dr Palmström), 30m (Dr Büchi), and 20m (Professor Broch). I hold that SSE cannot be criticised for following professional advice that erred on the side of caution.

[225]    Length BAM had major concerns about removing material from the debris pile, which sat on a steep slope. Any failure could harm anyone working in the tunnel. On being satisfied that it could remove material without creating instability, it did so. That shortened the length of the bypass tunnel to 605m.

 

Downstream Access Tunnel

[226]    BAM insisted on a downstream access tunnel as part of the recovery project. It yielded benefits in relation to ventilation, access, and safety. The parties also hoped that it would shorten the programme. Once BAM had removed the silt and some of the debris it could then construct the BPT from both ends. Otherwise BAM would have had to transport all the equipment, materials, and men from the reservoir end to the work site.

[227]    The DAT achieved some but not all of its aims. It allowed the quick removal of all of the silt and some of the debris. It also allowed the repair in the lower headrace tunnel to take place without interfering with the lining operations in the bypass tunnel. But it didn’t achieve the objective of driving the bypass tunnel from both ends. I hold that at the outset, it was a reasonable element to include in the design. 

 

Concrete Invert

[228]    The Hochtief experts thought that the installation of a concrete invert was not critical, despite it having been proposed at the Essen Workshop. The position of HSE closes any argument against the reasonableness of this part of the works. Lamont informed SSE in October 2009 that there should be a concrete invert. He would have recommended enforcement action if BAM had not done so. Apart from these safety considerations Hindle said that the concrete invert reduced construction time and cost.

 

Dam Bottom Outlet Culvert

[229]    Hochtief’s design incorporated a dam bottom culvert at the reservoir. It is a long concrete box (140m x 2.5m) situated in the base of the dam. During construction, it diverted the river past the dam works. During operation, it allowed the reservoir to be drained for maintenance and to address any safety issue. It also provided a ‘compensation flow’ of water to avoid detriment to other river users and aquatic life downstream. The compensation pipe lies in the roof of the culvert.

[230]    SSE drained the reservoir at the start of the recovery project and converted it back to a river diversion system. It did not, however, empty the secondary compensation pipe. The water within it froze, causing the pipe to burst and the roof of the conduit to fracture. SSE discovered the problem in May 2012. The severe cracking and spalling in the crown posed a risk to the integrity of the structure. Until it was repaired SSE could not fill the reservoir. BAM completed the repair works in time to recommission the scheme in August 2012.

[231]    I reject SSE’s contention that the Works Information should have contained an instruction to drain the compensation pipe. I prefer the opinion of Hochtief’s expert, Adrian Erwee. He said that SSE ought to have been aware of the risk of a pipe freezing if it was left full of water for three winters at high altitude in the Scottish Highlands. No‑one had expected the reinstatement of the river diversion system. SSE should have asked Hochtief for advice.

 

Three Concrete Plugs

[232]    Hunter suggested that the three large concrete plugs that BAM installed in the HRT should have taken less than half the time that was actually taken. When asked, however, about the details of their construction, he appeared not to recognise the scale of the work involved. I am satisfied that this was a reasonable cost to incur in respect of this element.

 

Plant & Equipment

[233]    BAM spent £100,000 a day on plant and equipment. I accept the evidence from the SSE witnesses that that level of spend was unsurprising. It was a very large civil engineering project. Tunnels place heavy demands on plant and equipment “an aggressive environment”. In my view it made good commercial sense for BAM to maintain three key items of plant: one in use, one in the workshop and one spare. That minimised disruption to the programme, which could have been very costly.

[234]    It was reasonable to have two camps for the 300 strong workforce. The men based at the lower site mainly worked in the DAT. Those at the upper site mainly worked in the BPT. If the men had all been based at the lower camp, it would added two hours to their working day to move them up and down the 11km haul road. Having two camps extended the amount of effective time in the tunnels. It was necessary to employ road maintenance crews to keep the access road clear in all conditions.

[235]    BAM’s crew sizes were not excessive. Its deployment of crews of 22 or 23 including 15 men inside the tunnel was reasonable. I do not accept Hunter’s view that a crew of 15 with 7 in the tunnel would have been sufficient. BAM had three crews working from each camp: a day shift, a night shift, and a five day rest shift. This was necessary to carry out an efficient programme, particularly when the requirement that the workers had up to date training certificates for all the different machines and were aware of all the British Standard, British Tunnelling Society safety and operating requirements.

[236]    BAM had to pay relatively high rates because it was difficult to recruit labour in the Highlands. It also lost specialist staff to the tunnelling activities that were being carried out at the time elsewhere with higher wage levels.

 

Secondary ‘Defects’

[237]    SSE did not press its claim in respect of the secondary tunnel features. It was right to adopt that approach. I do not accept that they constituted defects for which Hochtief was responsible. My reasons are as follows. First, I do not accept that defects were present in 2008 that the inspection parties either overlooked or ignored.

[238]    Second, SSE issued defects notice D058 on 24 March 2010. It alleged that there was insufficient tunnel support at 31 locations. Some days later two Hochtief employees entered the tunnel, but could not find the defects. The rock appeared to be of good quality at the specified locations. Another attempt to identify them took place on 13 April, when representatives of Hochtief, Donaldson, Jacobs, and BAM went into the tunnel with a photographer. The Hochtief representatives departed early after it became clear that the others could not identify the features listed in D058. Later it turned out that the chainages painted on the tunnel wall were incorrect. Using revised chainages, Becker was able to identify certain features referred to in D058, but they did not look like defects to him. Hochtief adopted a similar position in relation to three later defects notices issued in 2011 (D060, D072 and D073).

[239]    Third, Professor Broch inspected the HRT in January 2011. He concluded that the alleged defects “are all either rock features which could safely be left unsupported in an unlined hydropower tunnel, or minor rock fall outs to be expected on a first dewatering”.

[240]    Fourth, Billig analysed each of the alleged defects. Adopting a ‘worst case’ approach, he classified 28 as requiring no work, and 74 as requiring minor works at the first maintenance period. None in his view posed a risk to stability.

 

SSE Management Time

[241]    The parties agree that the sum of £1.3 million properly reflects the value of the lost management time that SSE incurred due to a number of its employees having to deal with the collapse and the remedial works.

 

Progress, Delays & Productivity

[242]    BAM delayed the start of the remedial works because of the very bad weather conditions in the winter of 2009/10. Dr Büchi, who lives in the Alps, said that BAM should have started sooner. I am satisfied, however, that the delay was unavoidable. The poor conditions affected setting up the site and commencing the works.

[243]    Despite the criticisms of Professor Broch, Dr Büchi and Dr Palmström, who all believed that the works were over-engineered and took too long, I broadly accept Hindle’s opinion that they were done in an “amazingly quick time” for the United Kingdom.

[244]    I qualify my view in one respect. BAM’s productivity in the first three months of the works was poor. Brand referred to that period as “the dark days of the project”. Fawcett noted that the work progressed “extremely slowly” (third report, 2 July 2010). On 10 August Appleby expressed concerns about the rate of progress and that the job was not being “progressed with the same vigour that one would expect of a contractor committed to a contract value and period.” Thereafter productivity improved as everyone became more familiar with the project. I would have been minded to shave a sum from any award made in favour of SSE to reflect the poor productivity in that quarter.

 

Adjudication Award, Insurance & Low Availability Damages
[245]    There are three other matters relating to loss.

[246]    Adjudication SSE maintains that, even if the collapse was an employer’s risk event, the sums of €339,356 and £28,248 that it paid on 6 July 2015 in line with the adjudicator’s award fall to be repaid as they were not costs incurred ‘due to’ the collapse. I would wish to be addressed more fully on this point before coming to a final view.

[247]    Insurance SSE’s claim under the CAR policy is the subject of an arbitration. I was informed that it has rejected an offer made by the insurers as being too low. Plainly SSE would not be entitled to recover any sums from Hochtief in respect of which it is indemnified by the insurers.

[248]    Low availability damages If the Glendoe station fails to achieve certain targets in the two year period after completion, Hochtief must pay ‘low availability damages’ to SSE in respect of loss of revenue: Clause Z6. That is subject to a £1 million cap which is, of course, vastly exceeded here. I am therefore minded to award SSE the sum of £1 million with interest. Hochtief contends that Z6 also bites in respect of the DAT, because it was constructed solely to reduce SSE’s loss of revenue. I would require further submissions before coming to a final view on this point. My provisional view is that the DAT was constructed for a variety of reasons and was an integral part of the recovery project. Accordingly it is not subject to the limitation cap.

 

22. The Counterclaim

[249]    Hochtief counterclaims for a sum of £5.4 million. It comprises the lost profit of £3.1 million that it says it would have made if it had carried out the recovery project, together with the costs that it has incurred and continues to incur as a result of the collapse. These include management time and payments to experts, lawyers and consultants.

[250]    Hochtief maintains that it was entitled to carry out the works. Accordingly SSE breached the contract by refusing to allow it to do so. Hochtief relies on Clause 83.1, which provides “Each Party indemnifies the other against claims, proceedings, compensation and costs due to an event which is at his risk”.

[251]    Although I hold that the collapse was an employer’s risk event, I also conclude that any loss suffered by Hochtief was caused by its own breach of the repairing obligation under Clause 82.1.

[252]    In any event from September 2009 onwards Hochtief’s relate primarily to monitoring the remedial works, and the costs associated with the adjudications and this litigation. They were not incurred due to the physical damage to the tunnel.

[253]    If I am wrong and Hochtief are entitled to recover in terms of the counterclaim, I note that there is a joint statement lodged in process where the parties’ experts have agreed that Hochtief has incurred and vouched the sums of €9,117,525.43 and £204,356.66. Sums of €97,004.22 and £217,977.85 have not been agreed and would require to be the subject of further submissions.

 

23. Observations on the Proof
[254]    I make the following observations at the end of this long proof.

[255]    Volume of Documents The parties lodged a total of over 73,000 documents. Combined with the daily transcripts of evidence, there was a truly daunting volume of material by the end of the case. It is unfortunate that the parties were unable to narrow the scope of the proof by entering into a joint minute of agreement. An agreed chronology and an agreed list of issues would also have been helpful.

[256]    Witness Statements (a) Contrary to the guidance of the Commercial Court, many of the witness statements appeared to have been carefully edited. Neither party made anything of this. I wish to emphasise, however, that the court derives the most benefit from statements which carry the author’s authentic ‘voice’. (b) Many of the statements rehearsed the facts, which resulted in a great degree of duplication. As each witness had a different perspective, all of them had to be carefully considered. (c) Some statements were in effect comprehensive reports. An example is the one provided by Dr Bodo Billig, Hochtief’s Head of Geotech. He and his two assistants took four months to prepare an 80 page statement containing many tables and technical information.

[257]    Adjudications I was not referred to the adjudicator’s decisions. Various witness statements and reports were, however, relied on in cross examination. Often this led to confusion as the nuances of different statements were compared.

[258]    Concurrent Evidence The most successful session of concurrent evidence involved the six tunnelling experts (Professor Broch, Dr Büchi, Professor Grøv, Dr Palmström, Professor Sloan, and Professor Stille) who were all present in court with Smith of BGS. I chaired their discussion and imposed time limits on their individual contributions. I found it an extremely valuable exercise and one which I would repeat in suitable future cases. Instead of hearing complex testimony weeks apart, I was able to hear the different opinions at one and the same time. They were also able to challenge one another’s position. This brought the topics into sharp focus. Each expert had to crystallise his position. I should add that they rose to the challenge by (in general) providing crisp answers. The exercise was less successful in respect of quantum, however, where there was little common ground and the level of detail was too great.

[259]    I am grateful for the considerable assistance I received from both parties’ legal teams. In particular I wish to record my thanks to senior counsel for the clarity with which they presented the case.

 

24. Conclusions
[260]    I summarise my main conclusions as follows:

a)   The contract imposed a duty on Hochtief to exercise reasonable skill and care in the construction of the scheme.

b)   Hochtief discharged that duty.

c)   The ground conditions were worse than it observed, however, and the support proved insufficient to prevent the collapse.

d)   Class  IV support was required to prevent the collapse.

e)   The collapse was not due to a defect that existed at take over. Accordingly, it was an employer's risk event.

f)    There was no contributory negligence. SSE was not at fault in respect of the odd readings and output swings from April to August 2009.

g)   Hochtief did breach its obligations by not returning to repair the tunnel.

h)   In consequence (i) SSE was entitled to instruct BAM to carry out the recovery project, and (ii) Hochtief is not entitled to recover any sums in terms of the counterclaim.

i)    If I had been awarding damages to SSE on its principal grounds of action, I would have held that the costs of the recovery project are reasonable, subject to deletion of the claims in respect of the ‘secondary tunnel defects’ and the dam bottom culvert. I would have also considered deducting a sum to reflect low productivity in the early months of the project.

j)    I am minded to award low availability damages to SSE.

[261]    I shall fix a By Order hearing to be addressed on the issues I have identified and to pronounce final orders.

APPENDIX

 

Map of location (southern tip of Loch Ness)

 
Source: Professor Broch's Expert Report [GLN0054699]: Figure 1

  

 

Source: Appendix 1 to Dr Büchi's Reply Expert Report [GLN0054767]: extract from 4minutes 32 seconds

 

Location photograph of reservoir / access road etc.

 

 

Source: Dr Smith's Presentation to the Court [Document 124 in process): Slide 5

Tunnel cross-section

 

 

Source: Dr Büchi's Expert Report GLN 0054765 : Figure 17.2 (annotation removed)

 

See also:  Photographs taken inside the Tunnel referred to below.

Tunnel Boring Machine

 

  Source: Dr Büchi's Expert Report [GLN 0054765] : Figure 31.1

 

 

 

Source: Dr Büchi's Expert Report GLN 0054765 : Figure 31.1

Conagleann Fault Zone

 

  

 

Source: Dr Smith's Presentation to the Court [Document 124 in process): Slide 7

 

 

 
 

Source: Dr Büchi's Expert Report GLN 0054765 : Figure 17.2 (annotation removed)

 

   

Source:  Headrace Tunnel, Downstream face of Blocked Zone, Post-collapse

 
  

Photographs of the upstream of the collapse dated 14 August 2009, 20 August 2009, 17 September 2009 and 10 November 2009.

 

 

Source: Source: Professor Broch's Expert Report [GLN0054699]: Figure 28

 

 

Source: Source: Professor Broch's Expert Report [GLN0054699]: Figure 29

 

Source: Professor Broch's Expert Report [GLN0054699]: Figure 30

 

 

 

Source: Source: Professor Broch's Expert Report [GLN0054699]: Figure 31

Needle Valves

 

 

Source: Mr Vinnogg's Expert Report [GLN: 0055028]: Figure 9

 

 
 

Source: Mr Vinnogg's Expert Report [GLN: 0055028]: Figure 11

REC Sheet – front and back

 

 

Source: Appendix 6 to Professor Broch's Reply Expert Report [GLN0054722]: (see separate document provided)