SSE GENERATION LIMITED AGAINST HOCHTIEF SOLUTIONS AG AND ANOTHER [2018] ScotCS CSIH

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	Scottish Court of Session Decisions
You are here: BAILII >> Databases >> Scottish Court of Session Decisions >> SSE GENERATION LIMITED AGAINST HOCHTIEF SOLUTIONS AG AND ANOTHER [2018] ScotCS CSIH_26 (10 April 2018) URL: http://www.bailii.org/scot/cases/ScotCS/2018/[2018]_CSIH_26.html Cite as: [2018] CSIH 26, 2018 SL 579, [2018] ScotCS CSIH_26

Pursuers and Reclaimers: Moynihan QC, Barne QC; CMS Cameron McKenna Nabarro Olswang

Defenders and Respondents: McBrearty QC, Richardson QC; Clyde & Co.

[1] This is an action of damages arising out of the collapse of a tunnel designed and

constructed as part of a hydro-electric scheme by the defenders for the pursuers at Glendoe,

Fort Augustus, between 2006 and 2008. At a basic level, the case is simply about which

party should bear the cost of repair; the employer who instructed the work or the contractor

who designed and built the tunnel. However, the litigation has proceeded along more

intricate lines. A major dispute has arisen about whether the collapse was as a result of a

defect in the design or construction. If it was, a further question is whether the selection of

an optional clause (M) in the standard form NEC Engineering and Construction Contract

(2nd ed), effectively converted what would otherwise be a contract requiring the contractor to

design and build something which was fit for its purpose into one which obliged the

contractor only to use reasonable skill and care in the design aspects of the work.

[2] The head race part of the tunnel (HRT) ran for a distance of 6.2kms from a reservoir

formed at the head of Glen Tarff, some 600 metres above the turbine. The tailrace tunnel

(TRT) is a further 1.9kms in length and discharges the water from the reservoir into Loch

Ness on its south eastern shore. A significant feature of the engineering of the HRT was that

it is built through the Conagleann Fault Zone (CFZ) at a depth of some 260m. The CFZ is a

tear in the ground caused by an ancient earthquake. It is an area in which rock conditions

[3] In about April 2009, not long after the completion of the tunnel, there was a major

collapse in the vicinity of the CFZ, about 2kms from the reservoir. This caused a complete

blockage over a distance of some 70m, with debris sloping down towards the turbine for a

further 600m. After considerable, and latterly cool, communings between the parties, the

defenders declined to carry out remedial works without being paid. The pursuers

instructed the Royal BAM group to construct a 605m bypass tunnel and associated works.

This ultimately cost about £137 million; a sum not far removed from the original contract’s

[4] There was a preliminary issue, which was resolved at a debate, on whether the claim

was relevant, standing the existence of, or a contractual provision for, a joint insurance

policy in respect of contractor’s risks (being risks other than those of the employer (infra)).

The commercial judge held ([2015] CSOH 92) that the claim was relevant notwithstanding

the existence of the insurance. He did so without the advantage of the various Opinions in

the UK Supreme Court in Gard Marine & Energy v China National Chartering Co (the “Ocean

Victory”) [2017] 1 WLR 1793. This issue is the subject of a cross appeal and will be revisited

[5] Much of the argument in the principal appeal centred upon the proper construction

of the contract; notably whether the collapse was a contractor’s risk. Such a risk included

loss or damage to the works which was due to “a defect which existed at take over”. Central

to that issue was whether the commercial judge was correct in his analysis of what

constituted a defect. He held ([2016] CSOH 177) that no defect existed, partly on the basis

that the defenders escaped liability because of the terms of Option M. These excluded

liability for defects due to the contractor’s design if he proved that he used reasonable skill

and care to ensure that his design complied with the “works information”.

[6] The pursuers contended that the contract provided that the defenders were obliged

to correct any defect, no matter what its cause might have been, at least up until the expiry

of the defects correction period, which occurred some two years after take over. The design,

it was said, had not complied with the works information, which, it was argued, specified a

tunnel with a design life of 75 years. The defenders had, in addition, not complied with part

of the design; notably a requirement to counteract the potential for the “erosion of erodible

rock” and to install support commensurate with specific classes of rock. The defenders had

not discharged the onus on them, in terms of the contract, to prove the use of reasonable

skill and care, especially given their failure to adduce testimony from their leading

engineering geologist, namely David Taylor, who had made the crucial decisions on tunnel

[7] The defenders adopted the reasoning of the commercial judge on his interpretation

of the contract and the discharge of the onus of proof. Although it had been accepted at the

proof that, as a generality, the defenders were liable to correct defects, the pursuers had not

demonstrated that a defect in the design had ever existed. At the centre of this argument

was the contention, which the judge accepted, that the design involved a holistic joint

process whereby the parties agreed the condition of the rock, and the measures required to

ensure adequate support for the tunnel, at the time of excavation and thereafter during a

[8] The reclaiming motion raised a number of subsidiary issues, including whether the

commercial judge’s reasoning was adequate and whether he could consider, and attach

significant weight to, documents (including reports), which had not been spoken to by their

authors. The judge’s award of almost all of the expenses, including those of an 84 day proof,

[9] The defenders cross appealed not only on the insurance issue, but also on the

commercial judge’s failure to award damages for the pursuers’ breach of contract in not

allowing them to carry out the remedial works. They contended that the judge erred: (i) in

awarding the pursuers damages on account of the scheme not being operable following the

collapse (low availability damages); (ii) in requiring the defenders to repay monies awarded

upon Adjudication; and (iii) in his assessment of quantum of hypothetical damages, had the

[10] Critical to the resolution of the litigation is a proper understanding of the structure of

the contract. In order to describe this, it is necessary to quote substantial parts of both the

contract and the associated works information and design documents. In terms of Schedule

Part 1 (Contract Data Part One), the contract’s conditions included the core clauses and inter

alia Option M of the standard form NEC Engineering and Construction Contract (2nd ed)

(reprint May 1998)1. The “works” were defined as those described in the “works

information” and all the documents incorporated in the contract’s Schedule Part 3. Under

the heading “2. Contractor’s Main Responsibilities” it was provided that:

“The contractor’s liability for defects due to his design that are not listed on the

defects certificate is limited so far as he proves he used reasonable skill and care to

[11] The completion date was initially set at 28 February 2009. The defects date was

104 weeks after completion. Notwithstanding that the limitation on liability for defects

under heading 2 applied only to those not included on the defects certificate, which was to

be issued before the defects date, the broader terms of Option M were then set out. These

“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”.

[12] The Core Clauses, which were in 9 parts, included the following:

1 Some of the words in the contract and the works information (infra) were capitalised or italicised.

14.1 The project manager’s or the supervisor’s acceptance of a

the contractor’s responsibility to provide the works or his liability for

works information requires to the project manager for acceptance. A

21.5 The contractor’s liability to the employer for defects due to his

design that are not listed on the defects certificate is limited to the

Possession of the whole site returns to the employer when the project

40.4 If a test or inspection shows that any work has a defect, the

43.1 The contractor corrects defects whether or not the supervisor

notifies him of them. The contractor corrects notified defects before

the end of the defect correction period. If the project manager and the

contractor agree before the end of the defect correction period, the

defect correction period is extended. The defect correction period

begins at completion for defects notified before completion and at the

43.2 The supervisor issues the defects certificate at the later of the

45.1 If the contractor has not corrected a notified defect within its

defect correction period, the project manager assesses the cost of

having the defect corrected by other people and the contractor pays

[13] There then followed certain bespoke clauses particular to this contract:

“46.1 (Clause Z2.2) The contractor’s obligations for the correction of

maintenance as a result of normal wear and tear or degradation in the

performance of the works as a result of the employer’s failure to

If the employer cannot use the whole or part of the works due to a

defect which arises after completion and before the defects date, the

defects date is extended by a period equal to the period during which

the employer, due to the defect, is unable to use the whole or part of

46.2 (Z2.3) If part of the works is repaired or replaced due to a

defect which arises after completion and before the defects date, the

defects date … is delayed by a period equal to the period between

completion and the date when the part has been repaired or replaced.

46.4 (Z3.1) Notwithstanding the issue of the defects certificate, the

possible speed any defect which appears or occurs during the period

of … twelve years following completion ... provided the defect is not

Upon receipt of written notice of such a defect the contractor without

receipt of the acceptance of the employer to the contractor’s proposals

… the contractor … implement his proposals as accepted with all due

If the contractor fails to submit his written proposals … the employer

necessary remedial work … and the contractor is liable for all costs

assesses the station availability monthly and notifies the contractor.

stated below at the end of each of the first four half year periods. The

availability standard and amounts for low availability damages are as

The total of the availability damages does not exceed £1,000,000,

which shall be the sole and final remedy for the event described in

80. Neither the contractor, nor the employer excludes or limits any

liability to each other for personal injury ... to the extent that such

injury results from the negligence or wilful default of itself, its

... the total liability of each of the contractor and the employer ... to the

other does not exceed the applicable financial limit for each category

... in the case of all liability for loss of or damage to the other party’s

property (except the works, plant and materials and equipment) ...

In the case of any other liability under the contract ... liability is

Neither the contractor nor the employer are liable to each other for

indirect or consequential damages and/or loss of profit except as

2 The insertion of the clause above resulted in there being two clause 80s.

81.1 From the starting date until the defects certificate has been

issued the risks which are not carried by the employer are carried by

82.1 Until the defects certificate has been issued ... the contractor

83.1 Each party indemnifies the other against claims ... due to an

83.2 The liability of each party to indemnify the other is reduced if

84.1 The contractor provides the insurance stated in the Insurance

provide cover for events which are at the contractor’s risk from the

85.2 Insurance policies include a waiver by the insurers of their

subrogation rights against directors and other employees of every

employer for events which are at his risk and by the contractor for

[15] It is worth observing that Option M appears in the following form, which includes a

“Option M: Limitation of the contractor’s liability for his design to reasonable

M1.1 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

[16] There were three relevant elements to the works information.

[17] First, part 2 of section 1 (“General”) of Schedule Part 3 contained a “description of

the works”. The description began (para 2.1) by setting out the employer’s objective of

having a hydro scheme designed and constructed, which would provide, inter alia: “Reliable

service without the requirement for major refurbishment or significant capital expenditure

within the design life of the scheme”. The works were (para 2.2) the design, construction

and commissioning of the scheme, which was to have a capacity to generate 100 MW.

[18] Part 6 of section 1 dealt with the “Contractor’s Design”. It provided (6.1.1) that the

contractor was responsible for the design of all of the works. It described (6.2.1) the design

procedures, which were to be carried out in three phases viz.: concept design; preliminary

design; and detailed design. Each element required (6.2.1, 6.2.8.1) to be submitted “for

acceptance” to the pursuers’ project manager, namely (initially) Neil Sandilands. The

detailed design, once accepted, was to be “developed through construction detailing and

method statements” (6.2.8.4). Design documents could be “accepted, not accepted (returned

[19] Part 6 of section 1 contained the “Design Brief” which provided:

The contractor designs a hydro scheme ... with a minimum guaranteed

performance. The performance criteria are detailed in the contract data

The design life of the principal components are detailed below. The

components provide reliable service without requirement for major

refurbishment or significant capital expenditure for a period of time as

[20] Part 7 stated (again, para 7.5) that the “defects date” was 104 weeks after the issue of

the completion certificate. The contractor was to attend a defects inspection during the final

[21] Secondly, section 2 (“civil works”) of Schedule Part 3 referred to the HRT. Part 2

dealt with general requirements including tunnel design; specifically (para 2.3.1) support.

The assessment of support to the HRT was initially to be based on the Q-system, which

involved an arithmetical assessment based on numerical values attributable to various rock

“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. The contractor agrees the rockmass parameters with the

project manager and the support class is agreed prior to its installation ...”.

The contractor could (as in due course the defenders did) adopt an alternative comparable

[22] Part 6 dealt specifically with the HRT. It was to be designed (para 6.1) to transport

water under pressure from the reservoir to the turbine. The initial concept was that the HRT

was to be lined throughout its length with shotcrete (sprayed concrete), except where a full

in situ concrete or steel lining was to be provided or as otherwise agreed with the project

manager. In due course, the proposal to have a completely lined tunnel, which formed the

[23] Thirdly, there was Schedule Part 7, Appendix 6, being the defenders’ “design

statement”. This had been prepared by the defenders’ subcontractors, namely Pöyry Energy

JV (Pöyry), a Finnish company which had been engaged to design the civil works, including

the HRT. It had been part of the tender documentation compiled in 2004. It provided an

explanation for the changes which had involved a major departure from the original lined

tunnel design. It assumed (para 2.2.2.1) the installation of shotcrete (75mm thickness over

50% of the tunnel with no concrete invert) where the rock was determined to be class I or II

(infra). It was anticipated that shotcrete would be required over 90% of the HRT. A lining

corresponding to classes III and IV (150mm shotcrete over the whole tunnel with a concrete

invert) was predicted over the remaining 10%. It was stated (para 2.4) by the defenders that,

with the use of a tunnel boring machine (TBM), the HRT was feasible with 60% of the tunnel

left unlined (class I). This, it was said, would not compromise its required service life.

[24] The design of rock support (subsequently called the Pöyry Guidelines) was to be

“Determine risk situations (geological scenarios) for which the rock support should

Design a range of typical sections (one for each excavation class) covering foreseeable

Determine forecast for intensity of risk situations along tunnel length and hence for

distribution of excavation classes along the tunnel [refer to reference ground

Determine actual support requirements during the tunnel advance based on actual

conditions encountered, particularly with reference to the identified risk situations;

Adapt rock support based on actual rock conditions (eg adjusting the balance of

shotcrete and rock bolts, adjusting the thickness of shotcrete, adjusting rock bolt

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

[25] The design statement provided (para 2.6) information to the effect that general

reinforcement was required when certain conditions were encountered. These included

areas where rock was susceptible to erosion and in areas where long term stability could be

[26] The works information contained a substantial report (Sch Pt 3; Appdx 12) setting

out the defenders’ understanding of the relevant ground conditions. These were said in the

report (para 1 “Introduction”) to form part of “the agreement for the works”. In relation to

the HRT, the excavation classes were said (para 4.1) to be defined on Drawing D2013

appended to, and in chapter 7, of the report. The Drawing was of cross sections of the HRT

showing the rock support applicable to each class. Chapter 7 began with a foreword which

“During construction, the support requirements will be decided at the face, based on

The “actual conditions” were to be understood as encompassing any observation likely to

affect the stability of the unsupported excavation. The class was not primarily a statement

of rock mass but something which reflected “the type and quantity of support installed

within the tunnel”. It was to be “considered as the adequate and sufficient response to a

[27] The report (para 7.4) divided the “excavation classes” into four. Class I (unlined)

was “good rock condition” with no hazards other than small unstable wedges or slabs of

less than (<) 0.2m3. Typical rock properties included an absence of sheared zones or erosion

potential. Either no support would be needed or it would be limited to spot bolting and/or

local shotcreting to the roof. Class II (partially lined) was “fair to poor rock conditions”

involving a hazard of, amongst other things, block fall due to unstable wedges or slabs.

Rock properties would encompass “blocky rock structure with closely spaced joints” and

single shear zones with a thickness <10cms and “non erodible gouge”. This needed

“systematic support above the spring line” (the half way point on the vertical). Class III

(fully lined) was “poor rock conditions” involving an unstable roof and side walls,

moderately to highly weathered rock, subject to ravelling (tangling), “weak erodible rock, eg

sheared mica schist and quartz mica schist with high mica content”. Rock properties

included sheared zones, notably “unfavourable orientation of weakness zones”. Systematic

[28] Class IV (fully lined with steel) was “very poor rock conditions” with unstable roof

and side walls, moderately to highly weathered rock, subject to ravelling. Hazards were

“fault zones with erodible kakirite (dip <500 or width >0.5m)” and “Loose erodible gouge of

faults”. Typical rock properties would be “faulted zones”, “very unfavourable orientation

of weakness zones”, typically occurring where there were fault zones with weak kakirite

and strongly sheared zones. Kakirite had earlier been defined as “Material resulting from

rock faulting/shearing. Structure of loose material, with no or low cohesion. Characteristic

of young faults”. Systematic support over the full perimeter was necessary, with shotcreting

and bolting being considered insufficient to stabilize the excavation.

[29] The report had anticipated (para 8.3.1.2) that the presence of “highly weathered rock

at tunnel elevation, apart from the fault and sheared zones was … unlikely”. Generally,

high strength rock was predicted. Hence, it was thought (para 4.1) that, over the whole

length of the HRT, only 1.5% would need class IV support, 6% would need class III, 23.8%

class II and 68.7% class I (cf Sch Pt 7, Appdx 6 supra). As matters transpired (infra), no

class IV support, and only a very small section of class III support, was installed.

[30] Two documents were said to constitute at least part of the detailed design. This

design was not part of the works information and its terms were therefore not part of the

formal contract, other than by reference. The first document was the HRT Rock Support

Methodology (RSM) (Drawing 152 DC 12 20 97 02 with attached comment sheet) which,

incidentally, showed (Fig 5)4 the CFZ at around chainage 2000 (the chainages being

references to the distance in metres from the reservoir rather than imperial measures). The

CFZ was thought to be some 115m wide. It was anticipated that it would contain

predominantly rock classes III (80%) and IV (20%) (see the text below Fig 5). Under a

heading “Water outflow & loss”, there was a statement (para 6.1.5) that “Erodible zones will

have to be shotcreted irrespective of the associated rock class”. Sometime after the

commencement of the excavation, the pursuers’ geology consultants, namely Jacobs (Jacobs

Gibb Ltd and Jacobs UK Ltd), had asked the defenders to clarify what was meant by

[31] Part of the RSM was Table 115. This was described as a “Geotechnical risk

assessment after excavation”. It contained a reference to the hazard of “erosion of erodible

rock during operation”. Operation meant the period during which water would flow

through the tunnel under high pressure. At the turbine, the pressure would be 60 bar,

although only about 20 bar at the CFZ. The hazard involved a “high” risk requiring

countermeasures in the form of the “application of shotcrete if not already

[32] The second document was the drawing D201, which had been referred to in the

ground conditions report (para 4.1 supra). It contained illustrations of a cross section of the

HRT showing what constituted the recommended support for each of the four rock classes.

Class I required two angled 2.2m rock bolts at, and 50mm shotcreting of, the crown. Mesh

was optional. Shotcreting could be reduced or omitted where good rock sections were

clearly identified. Class II involved four rock bolts at, and 50-80mm shotcreting of, the

crown. Mesh was compulsory. Shotcreting could be extended down the sidewall “as

required”. Class III showed eight rock bolts above the spring line, with 100mm shotcreting

throughout the circumference. The installation of a bolted steel plate to the crown was

optional. Class IV involved only two rock bolts but full circumference steel ribs at 1 – 1.25m

intervals. The 200mm shotcreting was also full circle with mesh. “Particular notes” on the

“1. Rock supports be installed to the extent required to meet the rock conditions

5. Rock support and surface treatment may be modified as considered

necessary to adapt to the actual geological site conditions at the work site.

7. Rock support methodology/materials specification see 152 DC 11 50 97 02”

[33] The contract was signed on 22 and 28 December 2005. At a meeting between the

parties on 2 February 2006, it was agreed that they would each employ geologists on site to

agree the rock classification as the excavation proceeded. A similar decision was minuted at

a meeting of 24 May 2006. This method was put into practice by means of rock excavation

classification (REC) sheets6 devised by the defenders’ principal engineering geologist,

namely Mr Taylor. The REC sheet template referred expressly to the descriptions of

classes I, II, III and IV (supra). The template anticipated completion by marking particular

portions of the text and ticking boxes representing the geologist’s view of potential hazards,

typical phenomena and typical rock properties. The geologist’s view of the overall support

category would be recorded in a box on the sheet. There was a small section into which

comments could be inserted. The sheets were signed as prepared by Mr Taylor, confirmed

by the defenders’ subcontractors, namely Pöyry, and approved by the pursuers’

subcontractors, namely Jacobs. The sheets covered a variety of different tunnel lengths.

Their content was converted onto maps, each covering a 25m section of tunnel. The

commercial judge found (para [64]) that, throughout the excavation, no-one sought to alter,

or otherwise query, Mr Taylor’s assessment of the appropriate rock class and support.

[34] The excavation of the main tunnel had begun with the TRT, which was completed in

February 2007. Thereafter the HRT was excavated with the TBM passing through the CFZ

in late August 2007. Although the defenders had expected to encounter poor rock

conditions in the CFZ (see supra), they did not notice any obvious signs of this. A minute of

a meeting on 9 September 2007 recorded by Angus Speirs, an engineer with Jacobs, noted

that indications of the CFZ were “imperceptible”. The requisite support was deemed to be

class I throughout almost all of the CFZ. Excavation of the HRT was completed in January

[35] At the proof, the pursuers focussed upon an REC sheet, dated August 2007, for

chainages 2117 to 21017 (ie in the vicinity of the CFZ). Although almost all of the rock class

descriptors ticked were class I or II, when it came to the general description of the rock mass,

it had been assessed as class III. The manuscript comments included:

“Some shear zones with QMS [quartz mica schist] kakerite (sic), local slab formation

& blockfall in sidewalls. Rotate CII support to LHS from CH2117-2107. Extend.

Shotcrete to maximum extents in shotcrete bay and backfill voids”.

The overall support category was given as CIIA; that is class II with an increased application

of shotcrete. On the equivalent mapping (for chainages 2125 to 2100)8, there was a drawing

of shears, some 1 to 1.5m apart, with blockfall having occurred prior to shotcreting. The

mapping showed a shear with a dip of <500 (it was 480). The rock mass description

contained a reference to follated rock with “closely jointed fresh QUARTZ MICA SCHIST,

mod weak to mod strong”. There was no note of kakirite on the mapping, despite the

[36] The REC sheet for chainages 2084 to 20829, which had been revised in February 2008,

contained eleven ticks referable to the descriptors. Four were for class I, six for class II and

one (general rock description) for class III. There was reference to the area being

“moderately sheared”. The revised classification was class I.

[37] An email from Mr Taylor to the TBM crew dated 28 August 2007 read as follows:

“Current conditions: Rock Class II. Now in anticipated fault zone so expect

There have been a few areas where there are local zones of better looking rock (a

couple of metres wide). Whilst these superficially look like Class I conditions, they

are bounded by Class II and may cover potential wedges formed in the weaker

ground. Whilst we are in this fault zone, it would be prudent to maintain a

minimum of Class II support unless the rock quality unexpectedly improves over

[38] In January and February 2008, inspections of the HRT took place prior to “watering

up”. Pöyry prepared a report recommending the application of strips of shotcrete and mesh

to certain areas of erodible rock. Defects notice D033 (infra), which was issued on 26 June

“A joint inspection of the [HRT] ... has identified areas where the rock support is not

in accordance with the agreed rock class and the support is considered to be

inadequate to ensure the long term stability of the rock. It is therefore not in

accordance with the Contractor’s design which has been accepted by the project

A number of specific items were said to require attention, notably additional support

measures, particularly shotcreting and extra bolting, in a significant number of locations

[39] On 11 March 2008, Prof Einar Broch, a geologist who had been engaged by the

defenders primarily to inspect the rock trap (a pit located in the HRT above the turbine),

reported that the rock mass conditions in the HRT were very good. No major weakness

zones had been observed. The zones described in the Pöyry report were small and none of

these could cause a serious collapse. Prof Broch regarded the support measures, which had

been recommended in the report, as being more than good enough for a tunnel that was

[40] Take over inspections occurred in October 2008. The pursuers had noticed certain

items lying in the invert. Mr Speirs had expressed a concern about the condition of the

lower side wall. He was reassured by Pöyry and the defenders that, if erosion occurred, it

would not adversely affect the stability of the tunnel. Defects notice D041 was issued on

10 November 2008 as a result of the joint pre-watering up inspection. This recorded that one

of the many defects, “whereby the construction was not in accordance with the Contractor’s

“3. Significant voids were evident below the tunnel invert at several locations

where weak and/or sheared rock had been eroded by the water flushing exercise

[41] Defects notice D044, dated 17 December 2008, again listed numerous areas of

outstanding work including filling “with concrete pits and voids in tunnel invert”.

Nevertheless, the pursuers issued the completion certificate to the defenders on

18 December 2008. This meant that the defects date would be 18 December 2010.

[42] Efficiency tests, which were required before the scheme could be connected to the

National Grid, were carried out in January 2009. Some time in about late May or early June,

and again on 10 July, 2009, odd readings began to appear at the pursuers’ control room in

Perth. These might have indicated a loss of water pressure at the turbine. They were

initially put down to calibration errors. It was accepted that, in hindsight, had they been

acted upon, matters might have turned out differently. The scheme was officially opened by

HM the Queen on 29 June 2009, when the turbine had initially failed to start. The following

day, thumping noises over a period of two hours were heard by Jacobs’ engineers. There

had been a progressive collapse of the tunnel from 12 April 2009.

[43] In early August 2009, judges from the Saltire Society were visiting. The parties

noticed a sediment plume of discoloured water being discharged from the TRT into Loch

Ness. The following day, the scheme did not achieve its operating load of 100 MW. It was

only at 80 MW after 25 minutes. The turbine was shut down. The pursuers arranged a

meeting at Perth on 6 August, by which time those involved had become “shell-shocked”.

Divers were sent down to inspect the tunnel at the reservoir end; but the problem was not

there. There was a note in the pursuers’ records stating “give direct to Hochtief now? – Neil

says no”. However, Neil (Mr Sandilands), informed the defenders of the problem later that

day, when defects notice D053 was issued. This was in the following terms:

“There has been an increase in headloss in the [HRT] ... resulting in [the pursuers]

not being able to run the generating plant at its rated 100 MW output and deciding to

shutdown the generating plant yesterday, 5 August 2009 ... in order to prevent

The support to the tunnel on part of its length would appear to be inadequate.

The [pursuers intend] to empty the tunnel to permit inspection and the project

manager will write to you giving details of arrangements and inviting you to attend

[44] Dewatering and inspection of the HRT took place. Rock falls could be heard. Minor

collapses had occurred, but the major one was discovered some 2kms from the reservoir.

An inspection from the other end of the HRT revealed some 600m of debris running up to a

complete blockage. It was initially thought that the blockage was some 270m long, but this

was later discovered to be erroneous; the final figure being some 70m.

[45] Although initial relations between the parties remained amicable, this soon changed.

By letter dated 31 August 2009, the pursuers, under reference to defects notice D053,

instructed the defenders to proceed with remedial works in terms of clause 43.1 of the

contract. The defenders refused to do so unless payment for the remedial works was to be

[46] By September 2009, it was accepted by both parties that the appropriate remedy was

the construction of a bypass tunnel. By letter dated 25 September 2009, the defenders stated

that they did not consider that they had any liability or obligation to carry out remedial

“We do however wish to emphasize that we are ready to repair the collapse as we

The defenders referred to clause 61.3; arguing that the collapse constituted loss of, or

damage to, the works taken over by the employer. Accordingly, the event was an

employer’s risk under clause 80.1 and a compensation event under 60.1(14). The cost of

repairs fell to be reimbursed under the employer’s property insurance. Under clause 83.1,

the defenders were entitled to be indemnified by the pursuers and compensated for all costs

due to the occurrence. The defenders said that there were no defects in the works noted at

the time of the take over and no evidence of any defect for which the defenders were

responsible. The parties’ positions have hardly changed ever since.

[47] On 27 October 2009, defects notice D055 was issued. This notified collapses and

block falls in the HRT from the reservoir to the blocked zone and continued:

“Ch 2150 – Total collapse, debris fills tunnel to crown; rendering tunnel inaccessible

Ch 2420-2735 – Massive in-tunnel debris, not filling the tunnel completely, from the

There were three references to voids and shears in the HRT in the area down towards the

turbine. The areas identified outwith the blocked zone would, in due course, form part of

[48] On 12 November 2009, the defenders offered to construct the bypass tunnel,

provided that this was on a cost reimbursable (Option E) basis. This was not accepted. In

late November 2009, the pursuers invited the Royal BAM group to tender for the bypass

tunnel. An estimate of £30 million was initially provided. On 7 December 2009, the

pursuers sent a letter to the defenders requiring them to provide a programme for the

remedial works and to agree to 50/50 cost sharing until liability could be determined. On

14 December 2009, the defenders declined this offer. On 18 January 2010, the pursuers

notified the defenders that they intended to appoint BAM to undertake the recovery project.

They formally appointed them to do so in March 2010. In January 2011, a supplementary

agreement was reached between the pursuers and the BAM, with a revised estimate of

£85.7 million. Meantime, the pursuers had issued defects notice D058 regarding secondary

defects. On 28 October 2011, the pursuers authorised a total of £137 million to be spent on

the remedial project. The works were completed in August 2012, when the scheme began

generating the required amount of energy; producing £20 million of annual revenue.

[49] On 7 June 2011 the defenders referred the issue of responsibility for the collapse of

the HRT to Adjudication. In a decision dated 24 November 2011, the adjudicator (Robert

Galbraith QC) addressed, first, the issue of the standard of care imposed on the defenders.

“... the Contract provides that for [the defenders] to escape liability [they] must

establish … that [they] ‘used reasonable skill and care’ to ensure that in choosing the

[50] The adjudicator considered, secondly, whether the defenders had used reasonable

skill and care. He noted (para 53.1) that it had been agreed (in the Works Information) that

the determination of the support requirements was to be based on the conditions

“That is how [the defenders] carried out the decisions as to rock support. Their

decisions, using the [defenders’] classification system, were being checked by Jacobs

(for the [pursuers]) using the Q system … I conclude that the tunnel support within

the collapsed area … was in accordance with the Works Information.”

The RSM was appropriate and had been properly applied on site. Accordingly, he also

“the tunnel support within the collapsed area … was in accordance with the

In an interesting passage, the adjudicator ended (para 56.13) his analysis of this issue as

“Of course, the contractual responsibility remained upon [the defenders], but there

seems little doubt that, as a matter of fact, the decisions taken at all material times

were decisions that commanded the support of all the professionals (including the

[pursuers]/Jacobs) on site. In such circumstances … [the defenders have] established

that [they] carried out [their] functions in mapping and producing the RECs, and

choosing the rock support categories, and confirming those in subsequent

[51] The adjudicator turned, thirdly, to whether the collapse had been a defect or a

“... the rock support works as designed and constructed did not constitute a Defect

for the purposes of Clause 11.2(15). It is true that a collapse subsequently occurred.

However, that was not because there was a Defect for the purposes of the Contract.

It occurred simply because there are inherent risks of rock falls and collapses in

unlined tunnels. These are part of the risks that the Parties are aware of when the

price is being negotiated. The incidence of such rock falls and collapses can be

greatly reduced by using a fully lined tunnel, but this would cost far more than an

He held that there was no defect which existed at take over “since the rock support specified

and installed was appropriate for the purposes of the contractual terms”. The damage,

which occurred after the works had been taken over, was an employer’s risk. The

adjudicator also held that the secondaries, being claims relative to areas of the tunnel

[52] A second adjudication proceeded before a different adjudicator (Richard

Fernyhough QC) who issued a decision on 26 June 2015. This ordered the pursuers to pay to

the defenders €339,356.31 and £28,248.81 in respect of costs incurred by the defenders in the

recovery project during the period from 5 August to 31 December 2009; all in terms of the

indemnity in clause 83.1, being costs “due to the collapse”. The costs were primarily those

of making the site safe, carrying out investigations, holding meetings and drawing up

outline remedial proposals (see para 46). The defenders had claimed much greater sums

amounting to €1,194,401.24 and £582,128.86, including legal and expert fees as well as their

own staff costs. A large proportion of those were disallowed on the basis that the causal

connection between the collapse and the costs had been broken when the defenders ceased

to be involved in the remedial project. Thereafter the work carried out by the defenders had

been as a consequence of commercial decisions, which were made in their own interests and

not effectively caused by the collapse, in order that they could defend themselves against

[53] Much of the proof was taken up with testimony about whether the defenders, and in

particular Mr Taylor, had been negligent in the assessment of the required support or in

their application of the RSM and drawing D201 criteria. In order to assess the commercial

judge’s findings in this area, it is necessary to delve some depth into the testimony of the

[54] Prof Sloan was the principal expert witness for the pursuers. He was a director of

Donaldson Associates. Prof Sloan visited the site on 14 August 2009, shortly after the

collapse, and again on 16 September 2009. In his report, dated August 2011, he concluded

that the collapse was centred on the CFZ; the presence of which had been “clearly”

referenced in the contract materials. The defenders had identified erodible rock in shear

zones within this fault, but they had failed to support and protect them in the manner

required by the works information and design. The defenders’ failure to install sufficient

[55] During operation, the action of the water in the HRT had eroded rock from multiple

shear zones. The erosion had been progressive and ultimately involved a large volume of

material, including blocks of rock that had been bounded by the shear zones. Upon

dewatering, a huge and catastrophic collapse of the ground above the tunnel had occurred.

A void of 2,374m3 existed to a height of 60m directly above the tunnel at the point of

[56] As a generality, site conditions had been suitable for an unlined tunnel, but careful

geological investigation and tunnel support had been required to ensure that the ground

conditions were recognised and dealt with appropriately. The RSM had not provided a

detailed explanation of how the excavation classification system ought to have been

implemented. No details had been given linking the collection of data to the selection of

support. The observations of actual conditions were to be given equal weight. The design

had not complied with the works information. It had assumed that the width of the CFZ

would be 5m, but this was contradicted by the information in the works information which

[57] There was a fundamental difference between rock classes I and II on the one hand,

and III and IV on the other; the latter having the potential to erode either immediately after

excavation or under operational conditions. Any rock described as having the attributes of

“erodible kakirite”, “weak erodible rock” ought to have been classified as either III or IV. To

be defined as III, any fault of shear zone had to be of a thickness of less than 0.5m wide with

a dip of <500. The only differentiating factors between classes III and IV was that, for

class IV, the width of any fault zone required to be >0.5m and the dip <500. In some 56 areas

the defenders had not, from their own data, identified the correct excavation class in

[58] There had been a significant amount of erodible rock in the collapsed zone and

frequent shear zones had been recorded over the full width of the CFZ. Although the

mapping had used class III terminology to describe these zones, the defenders had assigned

and installed only class I and II support. The works information and design required every

shear zone to be supported by the application of shotcrete over the full perimeter, as the

minimum class III support. The shear zones were not provided with class III or IV support,

because an assumption had wrongly been made that the performance of the rock would be

controlled by the dominant, and not the worst, recorded excavation class. In summary, it

had been apparent that the defenders had identified erodible rock that, based on their own

records, should have been attributed to class III. They had failed to do this.

[59] The February 2008 Pöyry design statement had recorded erodible rock at 11 out of

17 areas which they had inspected. It stated that the HRT, particularly the invert, had not

been inspected in detail. Twelve of the 17 areas referred to signs of deterioration,

bulging/squeezing, and shear zones. Two out of the 17 were referred to as “a big bad zone

with lots of weak material”. In four of the areas, over-break or wedge failures had been

identified. Many of the locations should have been identified as class III.

[60] The joint inspection of the HRT in October 2008 had noted unprotected sheared and

weathered rock that could erode under operational conditions. The pursuers had expressed

concerns about this. It had been explained that the agreed site approach was not to

shotcrete the lower sides of the tunnel walls. If erosion had occurred, it was said that it

would not significantly affect the stability of the tunnel. The defenders had therefore been

aware that there was rock which was susceptible to erosion but had been left unprotected.

[61] In the area of the collapse (chainages 2121 to 2050), the tunnel had been inadequately

supported. The requirements of the works information and the design had not been met.

Erodible rock was present and had eroded at multiple locations. Class III or IV support

should have, but had not, been installed. At chainages 2117 to 2101, there had been

evidence of faulting, using the criteria of erodible rock. The REC sheet had mentioned

kakirite and erodible rock over quite a length. This had not been reproduced in the

mapping. The REC sheet findings indicated class III. The assessment of this area as class II

had been caused by the absence on the mapping of erodible rock. The mapping had not

identified any need for class IV support. In parts it would have been class III. The extent of

erodible rock had been underestimated, albeit with hindsight. There was not enough

evidence to recommend class IV support, only class III. Erring on the side of caution meant

that class IV should have been installed, although class III would have stopped the erosion

and could have prevented the collapse. A collapse might still have occurred, but not one of

[62] Prof Stille had been instructed by the pursuers for the purposes of the litigation. He

had produced a report dated 22 October 2014. He had considered the RSM to be

appropriate for the expected geological conditions. There were ambiguities in the indicators

for defining classes in the RSM. The proposed support for the HRT in respect of the

different classes had been adequate, provided proper action was taken regarding the risk of

under-supporting the wall below the spring line and where there was erodible rock and

[63] In a second report dated 27 May 2015, Prof Stille stated that the contract had not left

responsibility for determining support to an engineering judgment at the tunnel face.

Rather, the support to be installed depended upon the actual hazard situation. If the hazard

was present, then the contractor had to classify and support the rock accordingly. The REC

sheet system had been adequate, but it contained ambiguities and had been susceptible to

subjective assessment. The worst case had to govern the selection of the level of rock

support. The assessment of an overall support category should have focused on the weakest

link. Although minor erodible zones and shears in otherwise good rock may not have

required shotcrete, the risk of instability due to erosion depended not only on the width of

the erodible zones, but on other factors, including the orientation and configuration of the

shears. The detailed design did not directly quantify the extent of erodible rock or gouge to

be considered as requiring support. The REC sheets had not done so either.

[64] The RSM had stated what was required when the rock had been classified as poor, or

very poor. In the absence of written instructions or method statements, the contract had to

be interpreted as meaning that all erodible rock, regardless of its thickness, should have

[65] The support for chainages 2117 to 2101 ought to have been class IV due to the

presence of the fault zone with kakirite, sheared joint surfaces and large over-breaks. The

HRT had been inadequately supported between chainages 2125 to 2050. The class identified

on the REC sheets did not accord with the design. The collapse had been caused by slaking;

that is to say the gradual disintegration of the rock when saturated. The appropriate action

would have been to cover all sheared mica schist, and quartz mica schist with high mica

content, with shotcrete either as class III or IV. This was what was stated in the RSM

[66] The Taylor email of 27 August 2007 suggested that the classification system,

including that on the REC sheets, had been for the purpose of identifying hazards during

excavation only. It was necessary to have two systems; one for excavation and one for

[67] At the time of the Pöyry inspections in February 2008, there had been a lot of

disintegration of material, especially in the invert. Steel plates had been put down;

suggesting that the bottom was not stable. If there were a problem with the bottom, there

was a problem with stability in general. The indication was that there was slaking, and the

risk of erosion. It should have been supported, as Pöyry had shown in the drawings, with

[68] Dr Palmström had been engaged by the defenders for the purposes of the litigation

as an expert geological engineer. He had produced a report which praised the use of the

Glendoe classification system, which (as distinct from the Q-system) had been created

specifically for the tunnel. The “design as you go” method was commonly used and had

great benefits in construction time and cost. At no point, in the area of the CFZ, had the

observable conditions merited class III or IV support. “No altered or crushed rock, shear or

fault zones with clay or clayish or earth-like materials (kakerite (sic)) are shown in the maps

or on the REC Sheets” in that area and “no instability was observed during the 14 months

before watering up” (para 7.2). Dr Palmström repeated this in a later “reply” report. In the

several inspections during this period, neither Mr Taylor nor Pöyry nor Jacobs had

[69] In relation to the reference to kakirite at chainages 2117 to 2101:

“… David Taylor did not … think that this occurrence of kakirite, ... was ... so

important ... he has … ticked out the class [III] or class [IV] in the individual

classification. ... [H]e has seen some clay in one or maybe more joints, but ... this clay

Dr Palmström accepted that his statement in his original report about the absence of kakirite

had been erroneous. It should have referred to “kakirite of importance”. Mr Taylor was an

experienced engineering geologist and, if he had seen kakirite of some importance, he

[70] An unlined tunnel should not be considered complete until it had undergone testing

by watering up and dewatering. After it had been operated for about a year, a tunnel

should be dewatered and inspected. A rock mass, which could erode through the effects of

running water, had to be protected sufficiently to eliminate the risk to stability. It would be

[71] Something must have caused the collapse. What it had been was unknown. The

collapse could have occurred in the manner described by Dr Rolf Wilhelm, viz.:

“What can be identified ... are however several smaller shear zones separated by

alternating layers of intensely folded weak mica schist layers, very strong quartz

schists, quartzite and thick bands of quartz. The local rock mass is therefore very

There were two shears on the mapping and the “moderately strong” quartz mica schist

required some explanation. Mr Taylor’s email of 28 August 2007 had recognised the

presence of the CFZ and that had been contradicted by the minute noted by Mr Speirs.

[72] The revised REC sheet for chainages 2084 to 2082 (supra) was “very strange”.

Table 11’s reference to “Erosion of erodible rock during operation”, requiring the

countermeasure “Application of shotcrete”, was a hazard during operation and required

“this is …the theoretical part of it, but we have also talked about judgment and

experience. So why it was not used in the collapsed zone must have been because ...

– if they saw signs of this, it was not judged at class [III]. ... [T]here were at least two

experienced engineering geologists, one from Jacobs and one from [the defenders],

looking at this section. There were geologists who had seen the conditions at the

surface, they knew that the [CFZ] was there. None of these people did ever discover

that you had conditions which could give a collapse. That’s the reason why they

… [T]he REC sheet is … a help to identify and make documentation of the things

they have observed, but in addition, ... more important, is the judgment done by ...

Even if certain features, which indicated rock with a propensity to erode, had been

discovered, what was being looked for were features large enough to cause a collapse.

[73] Material that was sheared and had a high mica content had the potential to cause a

collapse. If Pöyry had been asked to reconsider erosion potential, and they had not looked

at a particular area, this had been, using counsel’s words, a “missed opportunity”. If the

mapping was right, there had been no clear evidence that class III or IV support should have

[74] Prof Müller was the defenders’ senior design manager on hydropower projects until

2012. He had been design co-ordinator for the project. The detailed design had consisted of

the set of drawings and explanatory documents prepared by the sub-contractors, approved

by the pursuers and issued for construction. However, observation and selection, adaption

and adjustment of rock support were still required both during and after excavation. The

design process continued until the works were accepted as completed by the pursuers. The

agreed design procedure had included taking the six steps in the Pöyry design statement

[75] It was, in general, essential to assess how the rock mass would behave in water.

There was no doubt that the geotechnical technicians had properly assessed the rock mass

“on the spot”. Table 10 had referred to the excavation hazards in the TRT, whereas Table 11

had included hazards during empounding and operation. Table 11 was a risk handling

matrix used during the tender negotiations, listing every risk that might have come up. In

general, the risks of the HRT had been identified as low; otherwise the parties would not

have agreed to an unlined tunnel. Erosion had, in general, been considered to be a low risk,

but it might have been found in particular stretches. The rock’s stability due to hydraulic

loading would have been considered. Pöyry had included a paragraph in the RSM on water

inflow/outflow and potential erosion. They had been well aware of the purpose of the

[76] The Pöyry design statement of February 2008 had been created after Prof Müller had

left the site. The inspection team must have walked the length of the tunnel; otherwise they

could not have specified the sections for re-examination. The statement recorded that the

whole HRT, particularly the invert, had not been inspected in detail. That would have been

because of puddles and steel plating over the invert. The document indicated that they had

walked, or driven along, the whole tunnel, but certain areas were inspected in order to test

[77] Dr Büchi was an independent consulting geologist and tunnelling expert with his

own company, namely GEO 96, based in Switzerland. He had been instructed by the

defenders. In October 2014, he had produced a very long report covering many aspects of

the case. He considered that the defenders could not have predicted the actual geology in

the vicinity of the blocked zone from the information available at the time. The CFZ had not

been detected in the HRT. The defenders’ experience had been that class I or II support was

reasonable and appropriate, even in a predicted fault zone. The mapping of the conditions

encountered did not indicate a risk of a major collapse. The pursuers had confirmed the

adequacy of the mapping. The inspections had not shown any warning signals. The

[78] Dr Büchi commented in his witness statement that the situation as mapped at

chainages 2050 to 2120 was typical of what had been found generally in the tunnel. That for

chainages 2905 to 2825 was similar, with local overbreaks causing minor rock falls during or

[79] Dr Büchi’s initial report said that the collapsed area was only 8 to 10m long. The

bypass tunnel had been of excessive length. The majority of the secondary remedial work

had been unnecessary. Some minor repair and maintenance work after dewatering was to

be expected. In general, the secondaries were not defects and posed no threat to stability.

The remedial works had taken too long. The downstream access tunnel (DAT) had not been

required. The concrete invert had been inappropriate and constituted a change in the

project. The most reasonable solution would have been to tunnel through the blockage,

which would have shown a collapsed crown section of only 8 to 10m. Alternatively, a much

[80] Prof Broch had been instructed by the defenders. He had given evidence in the

Adjudication about whether the REC sheets had been used just for the purposes of

providing temporary rock support after excavation and not for the permanent operation.

This is mapping the situation in the tunnel during construction. Nothing more and

Knowing that the tunnel was going to be watered, a decision had to be made on the level of

permanent support, but that was a different stage. The level during excavation was

typically the contractor’s responsibility. Long term stability was typically the client’s or the

owner’s responsibility. These were two separate steps, which is why the tunnel had been

inspected and a report produced in October 2008. On several occasions, the HRT had been

inspected by the pursuers and their engineers, as well as the defenders, with a view to

[81] In evidence at the proof, Prof Broch repeated that construction was a matter for the

contractor, whereas operation was for the employer; at least in Norway. In February 2008,

the parties had been specifically considering erosion potential during operation. Further

REC sheets were prepared and some were amended. Prof Broch had inspected the HRT on

11 March 2008. He had reported on 30 April 2008 that the rock mass was very good; it being

one of the driest tunnels he had ever inspected. No major weakness zones had been

observed. Those in the Pöyry report had been small and none could cause serious collapse.

The support measures recommended by Pöyry had been more than good enough for a

[82] At chainages 2117 to 2101 there were records of sheared rock, mica, kakirite and

“... at what volume? How much? It has just been ticked off. It doesn’t mean that

there’s a lot of this or that. It only indicates that it has been observed. The

professional geologists there were responsible for making the necessary support.

They have come to a conclusion that the rock is safe enough or the rock mass in the

The Pöyry inspection in February 2008 had been the first after the completion of the

excavation. There was a record of erodible areas, but “that’s not erodible rock”. There were

eroded shear zones. He had suggested that some concrete might have been used to fill

them. The decision had been taken not to do anything before watering up, but to wait and

inspect it carefully at the first dewatering. This was a typical unlined tunnel approach.

Fallouts during dewatering is part of the unlined design. That is why there is a rock trap

and an agreed dewatering. There had been very careful mapping of the CFZ area. There

had been no indication whatsoever, as agreed by everyone involved, that anything was

going to happen. There was nothing in what had been exposed to indicate that there could

[83] In his report of May 2015, Prof Broch had expressed disagreement with the views of

Profs Sloan and Stille, in that it was his view that the design process continued right up until

the point of watering up. What the defenders had been doing during construction had been

not merely implementation, but carrying out a fundamental part of the design process. The

test at watering up and dewatering would inform the decisions about the sufficiency of

support. The performance of the tunnel would be reviewed. A risk had been taken in

February 2008 when some local problems were seen, but that was part of the construction of

an unlined tunnel. Any unlined tunnelling system necessarily had a greater residual risk of

rock fall than a fully lined system. An unlined tunnel was a trade-off between speed and

cost on the one hand and risk on the other. The employer accepted slightly more risk in

return for substantial cost and pound savings. The only way in which the risk could have

been eliminated would have been to line the tunnel fully with heavy support. Even then,

Hot Tubbing (concurrent evidence session involving Drs Martin Smith, Büchi and

[84] The session on whether the collapse could have been foreseen commenced with

Prof Grøv stating that an area of potential weak rock mass had been seen but “maybe not

appraised correctly” during the excavation and inspection phases. This did not mean that

the collapse could have been foreseen. If someone had foreseen a collapse, “they would

have done something more than what was actually done”. Many of the indicators for a

collapse had been observed and recorded. Prof Stille said that, theoretically, the collapse,

that is the slaking (disintegration of rock) or higher than anticipated load, could have been

foreseen. If lack of support could have been recognised, the collapse could have been

foreseen. Evidence of poor rock existed. The section of the tunnel in the CFZ had been

[85] Dr Palmström said that the mapping had been of good quality. It was easy to

understand what had been observed. The tunnel had been inspected several times. Areas

where additional support was required had been identified. A progress report had said that

the CFZ had been imperceptible. None of the experienced people involved had observed

any signs indicative of the potential for a “great collapse”. Dr Büchi said that, listening to

the argument about shears and slaking, the whole tunnel ought to have been lined. Those

observing the tunnel had been aware of a minor risk and had accepted that risk. The

[86] Prof Broch said that the collapse could have been foreseen, if the clear signals during

operation in April/May had been understood. A total collapse could have been prevented.

Dr Smith, who was the Science Director of the British Geological Survey, referred to signs of

the CFZ from the bore hole drilled prior to excavation. Prof Sloan thought that the

difference between the CFZ and the rest of the HRT had been identifiable during excavation.

[87] Prof Stille said that he understood Mr Taylor’s email to mean that there was unstable

ground at chainages 2101 to 2117. Pöyry had recommended a strip solution in certain areas.

There were mica schists, which would slake, connected to sheared mica schist. There was

some overbreak and some class II rock requiring additional support.

[88] Dr Büchi said that the rock mass in the CFZ was all class I or II. None would have

required III or IV. The overbreak was minor. The rock was weak and thus erodible because

of the presence of shears, which reduced cohesion and allowed water penetration.

Prof Sloan said that the presence of mica did not of itself mean that rock was erodible.

Prof Stille agreed with Prof Broch on cohesion and water penetration. Weak erodible rock

was defined in the contract as “sheared mica schist and quartz mica schist with high mica

content”. That type of rock would fall into pieces when submerged.

[89] Prof Broch did not know if mica schist, and quartz mica schist with a high mica

content, had been present at chainages 2050 and 2121. Dr Smith said that the mapping

demonstrated that it had been present. A shear had also been recorded. Prof Grøv agreed

and said that this had required class III support. Prof Stille also agreed; but whether the

support should have been class III depended upon whether the support related to

excavation or operation. Prof Sloan said that it had been clear from Mr Taylor’s email that

there was quartz mica schist in the CFZ area where the collapse had occurred. There had

been a reference to kakirite too. There was no doubt that the faulting and the influence of

[90] In relation to the required level of support to prevent the collapse, Dr Büchi said that

it was evident, from the shotcreting of the lower sidewalls and the invert, that what was

being considered had not just been safety but the final lining. Those mapping and

inspecting the tunnel had recognised the rock as erodible. The REC sheets had showed

erodible rock as class III, but this was only one of ten parameters. There was nothing to say

that a particular area had to have class III support. Dr Palmström agreed. It was a matter of

judgment. Prof Broch emphasised the combined nature of the mapping. It had been done

by a very experienced engineering geologist. It had involved a general evaluation of the

whole information available. The defenders, and the other “parties”, had been fully aware

that this was going to be a watered up tunnel and what was being talked about was long

[91] The commercial judge posed the question: do the experts agree that an engineering

judgment had to be exercised based on the actual conditions that were encountered?

Prof Broch and Drs Büchi and Palmström said “yes”. Prof Grøv said judgment had to be

exercised twice. Prof Stille said “yes, of course”, but the geologists had to follow the

contract. Prof Sloan agreed that judgments had to be made, but the design had mandated a

process of implementation by attributing the conditions to a specific class. Once that had

[92] In exercising judgment, Prof Sloan had seen no evidence of a rigorous joint mapping

exercise. Table 11 had set out the hazards during excavation and for long term stability.

The only consideration regarding the latter had been erosion. Prof Stille agreed. Table 11

was very important, where it stated that erodible rock had to be covered. There was no

judgment involved. Prof Grøv agreed with both. Dr Palmström said that the multiple

inspections had been carried out with a view to looking at possible slaking and erodible

rock. Dr Büchi said that, at the time of contracting, there were appreciable risks and that, if

detected, the parties had to act in accordance with the design. When erodible rock was

detected, it was at least class III, despite the presence all other factors. The mandatory

element came from para 6.1.5 of the RSM. Irrespective of class, erodible rock had to be

shotcreted. Prof Broch had been the only expert who had been in the tunnel before the

collapse. In March 2008 he, and Mr Taylor, had been taken to locations which, from a long

term perspective, could be dangerous based on the mapping of the previous February. They

[93] Turning to whether any of the experts felt able to criticise the judgments taken in the

tunnel, Prof Broch said “no”. Dr Büchi agreed that the rock was class I or II. Dr Palmström

recognised that it was difficult without seeing the actual conditions. He did not think that

he could, with hindsight, reach a conclusion on what had been done by experienced

geologists who had been working in the tunnel for a long time and had documented their

findings in the REC sheets and mapping. Prof Grøv said that the only clue from the collapse

zone was the mapping. He was unable to criticise what had been done.

[94] Prof Sloan took a different view. He did criticise the geologists on site. Having

walked through the TRT and HRT, he was of the view that there had been systematic

“under-realisation of the importance of erodible rock for the long term stability of this

tunnel”. There was a lack of attention to the consequence of leaving erodible rock exposed.

The defenders had been responsible for this. The erodible rock had been inadequately

[95] Prof Stille had criticised the overall classification system, but he would not criticise

the geologists. There were eight indicators relating to different issues, but no rule about

how the overall excavation class should be assessed, based on those indicators. That was

left to engineering judgment. This was a fault. If the contract had stated that erodible rock

had to be covered, implementation did not involve an engineering judgment.

[96] On the level of rock support required to prevent the collapse, Prof Sloan thought that

360° shotcreting would have gone a long way to prevent a progressive collapse. The

appropriate level of support ought to have been class IV (shotcreting with steel ribs).

Prof Stille thought that the erodible material ought to have been covered with shotcrete to

prevent slaking. This was not enough, given that there would be a poor bond with the rock.

A lot of rock bolts would also have been required. This is what Pöyry had described as a

strip solution. The invert had to be shotcreted too, as erosion at the invert could

progressively move upwards and lead to a collapse. The lower parts of the walls had to be

covered. Prof Grøv thought that the logical answer was class IV for chainages 2117 to 2100.

Class III would not have been sufficient. Dr Palmström did not think class II would have

been enough. Class IV was on the safe side. Dr Büchi said that class IV support would have

prevented the collapse. Prof Broch could not answer the question.

[97] In their summaries, Prof Broch said that the parties had been aware of the CFZ and

had carried out careful planning in that area. The defenders had been paid for the work

which they had carried out. There was no reason for them to have avoided installing the

appropriate support. Careful observation of the operation of the tunnel had been needed,

because there was some level of risk in having an unlined tunnel.

[98] Prof Sloan said that, when he visited the tunnel, there had been significant collapses

because of erodible rock within shear zones. The Pöyry design statement had contained a

mandatory requirement to protect erodible rock. Its roots were in Table 11, which looked at

both short and long (75 year) term support. There had been a systematic failure by the

[99] Dr Büchi said that the mapping had been agreed and there had been no mandate for

covering erodible material systematically with concrete. This had not been stipulated in the

contract and it was not what the parties had understood at the time. During mapping, there

had been no clear evidence of the presence of the CFZ. Several shear zones had been seen,

but these had fallen within class I or II. There had been no indication of a need for class III

[100] Prof Stille considered that there had been no systematic fault in the classification

system. There were some 20 areas in which the rock had deteriorated. These areas were

similar in type to the CFZ. The REC sheets did not cover the requirement in the RSM that all

erodible rock should be covered. The ambiguities in the REC sheets and the unclear

objectives were the reason for the erroneous decision not to cover weak erodible rock.

[101] Dr Palmström said that the engineering geologists had encountered the anticipated

stability problems in the CFZ. None of the rock had been class III or IV. Several inspections

had been made after excavation and dewatering. The tunnel had been tested with water for

almost a year. Some joints with erodible rock had been seen, but these were small features.

[102] Prof Grøv said that the rock classification system had its weaknesses. It had not

included certain important parameters and was missing a clear procedure on how to decide

rock class. Engineering judgment should only be used when the situation fell outwith the

[103] Dr Smith said that the CFZ could be seen on the surface and its existence had been

evident in the data sources. It would have presented a significant zone of weakened rock

where it intersected with the HRT between chainages 2050 and 2021. The mapping showed

distinctive features, such as increased quartz mica schist, intense foliation, fractures in the

form of shears, kakirite and overbreaks. The CFZ had caused the collapse. He felt unable to

comment on why the CFZ had not been recognised “for what it was”.

[104] The commercial judge concluded that the provision for joint insurance did not

displace the parties’ liabilities to each other under the contract. He accepted that the

authorities generally favoured joint insurance displacing contractual liability, but whether

that was so would depend upon the particular terms of the contract in their context.

Clause 83.1 expressly stipulated that each party undertook liabilities to each other. There

was no exclusion of liability (cf Co-operative Retail Services v Taylor Young Partnership [2002] 1

WLR 1419 (“CRS”)). There was no reason to justify overriding clause 83.1 and to give

primacy to the insurance clause 84. The requirement for a joint policy was based on

[105] If liability were excluded, clauses 83.1 and 80 (Z11 version) would be redundant.

There would be no need for provisions governing liability if there was no right to sue.

Clause 85.2 had waived the subrogation rights against the parties’ directors and employees,

but not in respect of the parties themselves. As was set out in Keating on NEC 3 (1st ed) (at

para 9-029), it is unlikely that clause 84 would displace a contractor’s liability for negligence.

Clause 85.4 maintained the underlying risk allocation. Clauses 83.1 and 2 implied that each

[106] In CRS (supra) it had been said that a stipulation for joint insurance carried with it an

implied term which prevented one party suing the other. There was no such implied term

in this case. It was not required for the purpose of business efficacy and it ran contrary to

the contractual terms. There was no irrebuttable presumption excluding liability simply

because joint insurance was in place. “That would tend to merge the law of insurance with

[107] The commercial judge heard evidence over some 6 months, primarily focused partly

on whether the defenders had been negligent in their performance of the contract, and hence

whether Option M might be applicable, and partly on quantum of damage. This focus was

reflected in the issues which the judge identified as being key. These were (Opinion

para [5]): why did the collapse occur; could it have been foreseen; were the defenders to

blame; were the pursuers also partly to blame; were the remedial works reasonable and

necessary? He classified the core of the dispute as lying in a deceptively simple issue

regarding whether the defenders had misclassified the rock and thereby failed to install the

[108] The short answer (para [147]) for the cause of the collapse was that “there was not

enough support: poor rock conditions coincided with insufficient shotcrete and rockbolts”.

There were formidable difficulties in providing a longer answer, because there had been no

full investigation of the cause of the collapse. It had ceased to be a priority once the

pursuers had decided to construct the bypass. The most likely explanation was as follows

“(1) The CFZ consists of interconnected faults of thin single shears with good rock

(2) The weak rock deteriorated and lost its strength when submerged ...

(5) The eroded material was progressively deposited as sediment over a

[109] The pursuers’ submission was summarised as being that a defect had existed at take

over and that therefore, in terms of clause 80.1, the risk was a contractor’s event. The

scheme could not provide the reliable service specified in clause 6.3.2 of the works

information. It did not conform to the accepted design, because the defenders had failed to

install the requisite level of support to prevent the erosion of erodible rock during operation.

“placed an important brake on liability. [The defenders] did not guarantee the

works. Instead, [they] accepted the familiar and lesser obligation of ‘reasonable skill

If the pursuers’ interpretation were correct, the defenders would have had an overarching

obligation to provide a tunnel suitable for its purpose. This would have imposed strict

liability and effectively robbed Option M of its meaning. It would also make some of the

[110] There had been no mandatory requirement to shotcrete all erodible rock (cf Table 11

“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.”

The contention that the defenders had failed to exercise reasonable skill and care was

rejected. The rock classification had been carried out jointly at the face. This was stipulated

in the works information, drawing D201, the ground conditions report and the design

statement. The parties had “clearly agreed the approach” (para [171]) whereby the

engineering geologists would jointly determine questions of both rock classification and

[111] Several factors were taken into account (para [174]) in determining that the defenders

had exercised reasonable skill and care, viz.: (1) mapping was a difficult exercise; (2) the

defenders’ experts had described the standard of mapping as of good or top quality; (3) the

TBM crew had said that Mr Taylor was an experienced engineering geologist who diligently

performed his role, took extra care at the CFZ, but had seen nothing. The fact that he had

noted weak rock on some REC sheets, reflected the care which he had taken. The judge was

satisfied with Mr Taylor’s judgment that no further support had been required. After all

“why go to the trouble of identifying such features and then decide to do nothing about

them?”; and (4) Jacobs and Pöyry had agreed with Mr Taylor’s classification. Many

experienced professionals had scrutinised the tunnel both during and after the excavation.

None of them had seen any faults that might have threatened stability. None of them had

[112] The defenders, Pöyry, Jacobs and the pursuers had all inspected the HRT on a

“metre to metre basis” in 2008, but had only uncovered minor issues, which had been

resolved prior to watering up. According to a report by David Fawcett, who was described

(para [181]) as a distinguished tunneller: “There is no recorded evidence of passing through

any feature that would potentially cause the catastrophic collapse that has occurred.” The

pursuers’ contention, that the REC sheets had been defective because they had only

addressed the hazards during excavation but not operation, was rejected. This had been a

“startling proposition”. After all, “why would [the defenders] construct a tunnel with no

thought to its long-term stability?” In short, the defenders had exercised reasonable skill

and care (para [187]). The pursuers’ case had been founded upon hindsight. Even if the

pursuers had established liability, they had not proved that class III support, for which they

were contending, would have been enough to avoid the collapse. The experts had agreed

[113] Clause 82.1 created a stand-alone regime. The defenders had breached this, from

August to December 2009, by failing to carry out remedial works. However, because the

collapse was an employer’s risk event, this had had a minimal impact on the pursuers’ loss.

They would still have required to bear the costs of the recovery project. It did mean, in

relation to the counterclaim, that, on the basis of the principle of mutuality, the defenders

[114] The pursuers had not been contributory negligent in failing to grasp the significance

of the odd readings from April 2009. The scheme had been a complex one and, even by

August 2009, it had not been in a steady (stable operational) state. Testing was still taking

place. It was understandable for the pursuers to have attributed rogue readings to

instrumentation errors. The signs had been confusing, with the scheme returning to full

output after each blip. The specialists, Andritz, had observed the odd readings, but had not

[115] The commercial judge adopted the general principle regarding action taken in an

emergency set out in Banco de Portugal v Waterlow & Sons 1932 AC 452 (at 506). The costs of

the recovery project had been “largely reasonable” (para [200]), taking into account

(para [202]) a number of factors, viz.: (1) the pursuers had had to act quickly to recover their

asset; (2) the pursuers had acted on the advice of professional consultants; (3) BAM, like the

defenders, had only been prepared to contract on the basis of reimbursable costs (Option E);

(4) the pursuers had instituted an exacting costs monitoring regime involving consultants

and auditors; (5) the remedial works had evolved over the course of the recovery project;

(6) no one had wished to put the scheme back into operation without being sure that it was

fit for purpose; and (7) it was not appropriate to compare the price of the original works

[116] A bypass tunnel (BPT) had been the correct solution, having regard to the

uncertainty about the size of the collapse, the extent of the void and the stability of the rock.

Tunnelling through would have carried a risk to safety. The BPT had not been placed too

far away, nor had it been too long. It had been reasonable to include a Downstream Access

Tunnel (DAT), because of the benefits it yielded in relation to ventilation, access and safety.

The Health and Safety Executive had insisted on the installation of a concrete invert. This, in

any event, had reduced construction time and cost. Although it had not achieved its object

of allowing the BPT to be excavated from both ends, it had allowed for the quick removal of

silt and debris and permitted repair of the secondaries without interference to the BPT

[117] The defenders had relied on John Hunter to demonstrate the excessive nature of the

remedial works, which he had estimated ought to have cost only £30.5 million. However,

the commercial judge rejected his evidence as Mr Hunter had, in contrast to the pursuers’

experts, failed to appreciate fully the initial uncertainty, the scale and sequencing of the

works and the rigour of the pursuers’ cost monitoring. Overall, BAM had made a profit of

only 2 to 3%. Mr Hunter had accepted that he had placed excessive reliance on the annual

costs reference work (Spon’s) and on a comparison with a nearby tunnelling project.

[118] The pursuers had not pressed their claim in respect of defects in areas of the tunnel

unaffected by the collapse. It was rejected in any event for 4 reasons. First, no such defects

had been noticed in 2008. Secondly, after the pursuers had issued a defects notice (D058) in

March 2010, which had stated that there was insufficient tunnel support at 31 locations,

attempts to identify these in April 2010, by the defenders, Donaldsons, who were the

pursuers’ consultants on the recovery project, Jacobs and BAM, had all been unsuccessful,

although it later transpired that this was because the chainages painted on the tunnel had

been incorrect. Once the features had been discovered, the view of the defenders’ senior

design manager, Thomas Becker, was that they did not look like defects. Thirdly, Prof Broch

had inspected the tunnel in January 2011. He had taken the view that what were alleged to

be defects were either rock features, which could be left unsupported, or minor rock falls,

which were to be expected on a first dewatering. Dr Bodo Billig, the defenders’ head of

geotechnics, had looked at each defect and had identified 28 as requiring no work and 74

requiring minor works at the first maintenance period. None posed a risk to stability.

[119] The recovery work had been carried out within a reasonable time, other than in

relation to the first three months, when productivity had been poor. The report by

Mr Fawcett had described the work as progressing “extremely slowly”. A deduction

required to be made to reflect this poor productivity (see infra).

[120] In terms of the contract, if the scheme failed to achieve certain targets in the two year

period after completion, the defenders were bound to pay “low availability damages” to the

pursuers for loss of revenue (clause 48.2, Z6). This was subject to a £1 million cap. This

amount was awarded to the pursuers, although further submissions were required on

whether the £1 million cap applied to the construction of the DAT.

[121] The defenders’ counterclaimed for £9.6 million. This consisted, first, of loss of profit

of £3.1 million, which the defenders would have made, had they carried out the recovery

project under clause 82.1 or, in response to the defects notices, clause 43.1. Secondly, there

was a claim for £6.5 million, being the costs which they averred that they had incurred in

terms of management time and payments to experts, lawyers and consultants in the

monitoring of the repair works, the conduct of the adjudications and this litigation. The

commercial judge held that any loss of profit suffered by the defenders had been caused by

their own breach of the repairing obligation under clause 82.1. Any costs incurred were not

[122] The commercial judge summarised (para [260]) his conclusions as follows:

“(a) The contract imposed a duty on [the defenders] to exercise reasonable skill

however, and the support proved insufficient to prevent the collapse.

(e) The collapse was not due to a defect that existed at take over. Accordingly, it

(f) There was no contributory negligence. [The pursuers were] not at fault in

respect of the odd readings and the output swings from April to August 2009.

(g) [The defenders] breached their obligations by not returning to repair the

(h) In consequence, (i) [the pursuers] were entitled to instruct BAM to carry out

the recovery project, and (ii) [the defenders] were not entitled to recover any sums in

(i) ... the costs of the recovery project [were] reasonable, subject to a deletion of

the claims in respect of the “secondary tunnel defects” and the dam bottom culvert

[frozen pipe]. [A sum ought to be deducted] to reflect low productivity in the early

(j) [Low availability damages should be awarded to the pursuers].”

[123] Having heard the parties further, the commercial judge held that the defenders were

obliged to reimburse the €388,720.27 and £32,357.98, which had been paid in terms of the

adjudication awards. The capped amount of £1 million for low availability damages would

be awarded in respect of the defenders’ breach of clause 82.1. Had the pursuers succeeded,

£107,617,830.94 would have been awarded in respect of the recovery project. This consisted

of £129,818,563.94 being the sum paid to BAM, plus the pursuers’ management time of

£1,338,267.76, less deductions of: (a) £533,345.76 in respect of low productivity (although this

sum was precisely that for management costs requiring to be deducted from BAM’s account

according to Mr Hunter); (b) £1,005,655 in respect of disallowed culvert (frozen pipe)

repairs; and (c) £22 million in respect of the cost of repairs to the secondaries. Interest at the

rate of 4% per annum from the date of citation was awarded as the claim was one for

[124] The defenders were entitled to the expenses of the action, because they had

substantially “vindicated their position”. A deduction of 5% was made to reflect the success

of the pursuers in respect of the adjudication awards, the low availability damages and their

successful resistance of the counterclaim. Time had been taken up enquiring into a matter

where there had been a lack of candour on the part of the defenders in relation to their true

[125] The pursuers adopted their written Note of Argument, without exception, but

presented an oral submission with a markedly different content. This summary attempts to

consolidate the grounds of appeal, the Note and the oral submission. In the Note, the

pursuers’ claim was based on three primary grounds: (i) a failure to correct defects

(clauses 43, 45 and 46); (ii) a failure to replace or repair loss and damage (clause 82.1); and

(iii) an indemnity for a contractor’s risk (clause 83.1). The parties were agreed on record that

the defenders had been contractually obliged to correct defects. If there was a defect, the

[126] In their oral submission, the pursuers’ central contention was that, despite the

presence of several “rabbit holes”, there was a clear route governing liability. The pursuers

had alternative cases. The first (ground of appeal 1) was that the commercial judge had

erred in his approach to the provisions on defect correction, by regarding Option M as

overarching. Secondly (ground 8), if Option M operated as a brake on liability, the judge

had erred in holding that the defenders had discharged the onus of proof.

[127] The facts could be stated in eight propositions: (1) the tunnel had been procured to

provide reliable service for 75 years; (2) the collapse had been due to a failure to provide

enough support and had been caused by erosion, which was a hazard highlighted in

Table 11; (3) there was no evidence that those who had designed and installed the support

had considered Table 11; (4) erosion had been observed prior to watering up, but had been

carried over to the first maintenance inspection; (5) there was no evidence that measures to

counter that erosion had been considered; (6) the obligation on the defenders had been to

provide the works. The defenders had an obligation to inspect and to correct any defects at

their own cost for at least two years after take over (completion); (7) the defenders had

refused to correct them at their own cost; and (8) given that breach, the defenders were liable

to pay damages of at least £107 million, or £126 million if the pursuers’ argument about the

secondaries were correct. If the court agreed with these eight propositions, liability followed

and there was no need to consider the rabbit holes of: (i) causation; (ii) contributory

[128] The contractual structure could be described in five propositions: (1) the defenders

were obliged to correct all defects; (2) the source of the obligation was in clauses 20.1 and

43.1; (3) there was an express provision that, if the defenders failed to correct a defect, the

pursuers could engage another to do so and recover the costs (clauses 45.1 and 46.4); (4) the

obligation under clause 43.1 subsisted until the defects date, which was two years after

completion. There was no distinction in the contract between the position prior to take over,

in December 2008, and the ensuing two years. If the tunnel, on inspection within the two

year period, was not performing satisfactorily, it was for the defenders to remedy that, at

their cost; and (5) the defenders had used Option M not just as a limitation on liability, but

as a means to create a right to carry out the work and to be paid for it.

[129] The commercial judge erred (at para [260(a)]) in his interpretation of the contract,

and in particular Option M. Clauses 20.1, 40.4, 43.1, 45.1, 46.4, 60.1 and 80.1 obliged the

defenders, at their expense, to correct any defect, irrespective of whether it had been caused

by a failure to exercise reasonable skill and care. The judge had not considered clause 20 or

section 4. He had found that there was an obligation on the defenders only to exercise

reasonable skill and care in the design of the scheme, but that was not what the contract

said. The proper starting point was to ask whether there was a defect. If the answer to that

was no, then whether reasonable skill and care had been exercised was irrelevant. The

pursuers attached to Option M the same interpretation as was given in the Guidance Notes

[130] The contract was arranged in such a way that the actions of the parties were defined

precisely (Guidance Notes, “Clarity and Simplicity”). The basic obligation on the contractor

was to provide the works (clause 20.1). This included testing and making good defects

(Guidance Notes, on clause 20.1). The issue of the defects certificate signified the end of

most of the obligations. There was unlimited liability for defects specified in the certificate;

ie those identified in the first two years. There was then a limitation on liability for defects

which had not been identified; ie latent defects emerging after two years. The Guidance

Notes said (para 21.5) that the contractor’s liability could be limited to an amount stated in

the contract data or by the inclusion of Option M. Option M was thus dealing with a

limitation on compensation. The limitation applied only after the issue of the defects

certificate (ibid, under reference to clause 43.2). The contractor’s liability to compensate,

rather than to correct, applied only after the issue of the certificate.

[131] Loss and damage occurring after take over, except in so far as relating to a prior

existing defect, was an employer’s risk. The point of take over was significant in terms of

section 8, which dealt with indemnity and insurance; but take over had no relevance to

section 4, because the obligation to correct was the same before and after take over.

Clause 40.4 provided that, if a test or inspection showed a defect, the contractor had to

correct it and repeat the test. The defects date was 104 weeks after completion. The

contractor had to attend an inspection within the final four weeks.

[132] On Option M: (1) there was a distinction between compensating and providing or

completing the works; (2) Option M qualified the liability to pay compensation, but not the

obligation to provide or to complete the works; (3) the contract spelled out the actions which

were required and their financial consequences. In the period down to the defects date, if

the contractor did not do the work, the costs could be recovered from him (clause 45.1).

Clause 46.4 extended the period to 12 years; (4) recognising Option M as a defence to a

failure to correct defects was to turn the provisions in section 4 on their head; (5) the

pursuers’ meaning was the one given in the Guidance Notes and was consistent with the

pragmatic analysis in CRS (supra); (6) where there was to be additional payment for work,

the contract said so expressly (clause 60.1). It was a shield and not a sword. It was a defence

to liability and not an entitlement to payment; and (7) there were express extensions to the

[133] The commercial judge erred (para [260(e)]) in holding that a defect had not existed at

take over. His findings (paras [147] and [152]) showed that there had been a defect, as

defined in clause 11.2(15), at take over. On 6 August 2009, the pursuers had issued defects

notice D053, stating that the support to the tunnel appeared to be inadequate. Eight years

later, the Lord Ordinary concluded that that was correct. An instruction had been issued on

31 August 2009, asking the defenders to proceed with the remedial works. They refused

[134] The commercial judge had addressed section 8, by asking whether there had been a

defect at take over. He had not addressed the route 1 argument based on clause 20.1 and

section 4. The judge held that the brake, that was engaged by Option M, meant that there

was no defect in the first place. The correct approach was to ask: first, whether there was

any defect at all; secondly, when it came into existence; and, thirdly, how the contract

addressed the defect? In terms of clauses 80.1 and 81.1, loss and damage occurring before

the defects certificate, which was due to a defect which existed at take over, was a

[135] The commercial judge found that the cause of the collapse was the absence of

adequate support. The HRT was therefore not going to give reliable service. There was a

failure to meet that standard. Similarly, and more specifically, the erosion and the slaking

meant that there was potential instability, and for that reason also the HRT was not going to

provide reliable service. The HRT thus failed to meet the requirements of the works

information. Insufficiency of support, as at take over on 18 December 2008, was a defect. If

the collapse was classified as loss and damage, as distinct from a defect, it was still a

[136] The parties were agreed that the contract between them was a fitness for purpose

contract with an exception; that being Option M. Option M was a limitation of liability

clause. It did not reduce the scope of the work or alter its nature. During the period up to

the defects date, the employer could ascertain whether the design was adequate for the

design period. For an unlined tunnel, that required the employer to put the tunnel into

operation in order to try it out. When the tunnel had been subjected to testing on a trial and

[137] At take over, there was an expectation that the outstanding defects notice would be

dealt with under clause 43. Defects notice D041 said that voids in the invert were to be filled

with concrete and remedial action was to be taken during the first dewatering. This

indicated that there was a section 4 defect at take over, namely that erodible material had

not been protected, in breach of the contractor’s design. It was also a breach of the works

[138] The requirement to design a tunnel to last for a period of 75 years had not been

qualified by Schedule Part 7. The defenders had not led any evidence that some other

period had been agreed. If the pursuers had inspected the tunnel in May 2010, and

observed cracking which was consistent with a threat to stability, they would have

concluded that the tunnel would not last 75 years. The question was not whether, in the

abstract, the tunnel had been designed to last 75 years, but rather whether at the point of the

[139] If the support in the tunnel required to be upgraded (ie class III rather than II) the

contractor would be paid for the increase in the level of support. The parties anticipated

periodic payments up until, and beyond, the completion date; the long stop being the

defects date in December 2010. A change in the percentage length of each class of support

would constitute a compensation event under clause 60.1(12), regardless of when the

variation occurred or the circumstances giving rise to the variation. The contractor was not

paid to correct a defect. An increase in the level of support was not regarded as the defect,

[140] The pre-watering up inspections were an event within the meaning of clause 80.1.

The collapse was due to the failure properly to address the risk of erosion that was observed

in certain sections of the tunnel. The collapse was accordingly a contractor’s risk. The

[141] The Lord Ordinary had given Option M too wide an interpretation. It only applied

to defects caused by the failure of the contractor’s design. It did not apply to a defect

attributable to a failure to implement that design. The design, if properly implemented,

would have resulted in the installation of class IV support. There had been a failure in that

regard. A failure to comply with Table 11 was not a defect in the design. The design

[142] The commercial judge had erred (paras [157-8]) in holding that the defenders had

discharged the onus of proof stipulated in Option M. The defenders’ chief decision-maker,

Mr Taylor, had not given evidence. The defenders had lost contact with Mr Taylor. The

pursuers had been in contact with him, but he could not speak to them because of a

confidentiality agreement. A standoff had ensued. Several others, who had been involved

in the tunnel, had not given evidence. Jacobs’ geologists had not been called, nor had their

contract supervisors. The author of Pöyry’s RSM had not given evidence, nor had those

who had signed the REC sheets. None of those who had taken any of the decisions on the

level of support had been called. Instead, a number of experts had been asked to look at

documents, with a view to telling the court what their authors had seen, thought or done.

[143] Although, when all the evidence was out, onus seldom mattered (Salt International v

Scottish Ministers 2016 SLT 82 at para [45]), there remained situations in which it was a

determining factor (AW v Greater Glasgow Health Board [2017] CSIH 58 at para [57];

Armstrong v Glasgow City Council [2017] CSIH 56; Stephens v Cannon [2005] CP Rep 31, under

reference to Rhesa Shipping Co v Edmunds (“The Popi M”) [1985] 1 WLR 948; McEwan v

[144] It was not known what those, who had made the relevant entries in the REC sheets,

had understood by “erosion”. It was not known whether they had considered Table 11. The

REC sheets, which had been used for the HRT and the TRT, had been identical to one

another, except for the header and footer. It was not known what the individuals looking at

the REC sheets had been thinking about when they had been taking decisions in the HRT as

[145] There were two shears with a gradient of <500 marked on the mapping for one

section (chainages 2125-2100) of the blocked zone (2121-2050). In terms of the RSM, this

required class III or IV support. On the equivalent REC sheet (2117-2101), there was a record

of kakirite, which also required class IV support. It was not known why class III and IV

support had not been installed. In the absence of the decision takers, the court could not

know what factors had been considered (Hills v Niksun Inc [2016] IRLR 715).

[146] The commercial judge made a significant number of errors of fact. He stated

(paras [180]-181]) that none of the experienced tunnelers had seen any signs of a fault that

might have threatened tunnel stability. None had recommended the installation of a higher

level of support. That was wrong. At chainages 2990 to 2895, which had been described as

the “big bad zone”, multiple shears had been noted. Pöyry had raised the need to apply

shotcrete to erodible zones, independent of the rock classification. In January 2008 they had

proposed that they return to confirm that the design requirements had been understood.

They had recommended the installation of 3600 support in areas (chainages 2117-2107)

where erosion had been observed. The judge found that Pöyry had inspected the tunnel on

a metre by metre basis in 2008. In fact, only certain parts, and not the CFZ, had been

inspected. They had not examined the collapsed zone (2121 to 2050). The part of the tunnel,

where additional support had been required, involved rock conditions of a similar nature to

the collapsed area. The failure to inspect the CFZ had been, in the words of Dr Palmström, a

[147] The commercial judge, in stating that there had only been minor issues, which had

been resolved prior to watering up, had been thinking of defects notice D033 in June 2008.

He had forgotten about defects notice D041 of 10 November 2008. In that, the pursuers had

recorded significant voids below the tunnel invert, where weak and/or sheared rock had

been eroded. There was erodible rock at a number of locations on the lower side wall. The

pursuers had been concerned about material in the lower part of the tunnel deteriorating.

The decision to leave this section exposed had been inconsistent with the Pöyry

[148] The critical words in the RSM were the references to fault zones with erodible

kakirite and a shear with a dip of <500 angle to the horizontal. The mapping (chainages 2117

to 2101) had showed a shear of <500; so the area required class IV support. The REC sheets

showed the presence of kakirite and shear zones with quartz mica schist. The overall

support category was inexplicably assessed at class IIA. Those on the ground may not have

complied with Table 11 because the REC sheet template did not remind them to do so.

There was a prima facie case, based on a combination of the mapping and the REC sheets,

[149] The parties had originally believed that the CFZ was between chainages 2075 and

1960 and not between 2117 and 2101. The email from Mr Taylor, dated 24 August 2007, had

suggested that the CFZ was between chainages 2070 and 1960. The TBM crew engineer had

received an email from Jacobs dated 18 July 2007, stating that the CFZ was anticipated

between chainages 1850 and 2200. This differed from Mr Taylor’s email. Both Mr Taylor

and Jacobs had been signatories to the REC sheet for chainages 2117 to 2101. There was a

conflict about where the CFZ was. The RSM had depicted the CFZ at chainage 2000. It had

been anticipated that there would be large continuous areas of disturbed rock, but this was

not what was noted. Rather, there were veins of faulting, interspersed with good rock. That

was the pattern which Dr Wilhelm had ultimately recognised. Table 11 contained the

reference to the hazard of erosion during operation. This criterion had not been contained

within the REC sheets. If the sheets were to be the sole guide to long-term support, they

ought to have contained all the criteria, including those in Table 11. The commercial judge

had made no finding that any engineer had considered the content of Table 11 as part of any

[150] The defenders had failed to prove reasonable skill and care because: (1) the REC

sheets for the HRT did not refer to Table 11; (2) no witnesses mentioned Table 11 as a

consideration; (3) the defenders had put forward contradictory positions in relation to

Table 11, including the proposition that it did not relate to the protection of erodible rock;

and (4) the REC sheets dealt only with excavation and not the second phase of operation.

[151] According to Prof Broch at the Adjudication, the REC sheets simply identified the

support to be used when the tunnel was being excavated. After that, someone would come

along and make a decision about the long-term support. If that were so, Table 11 was not

relevant to the first phase of the work. The commercial judge considered only what

individuals may have seen. He overlooked the evidence that the TBM crew and others may

[152] The commercial judge had failed to take into account that there was a two stage

process at which the REC sheets were completed. There was the initial excavation, followed

by a consideration of operational matters. Prof Stille noted that it would not be unusual to

take one approach to the excavation of the tunnel and another to the permanent support.

The RSM considered rock stability during excavation. It did not consider stability in

conditions of hydraulic load. That was why Pöyry had returned to inspect the tunnel and

consider erosion. In the absence of evidence that Table 11 and the erosion in the invert had

[153] Jacobs had specifically asked the defenders to clarify what had been meant by

“erodible” in the RSM, but the defenders had not done so. The commercial judge had failed

to conclude that constructing tunnels, in the absence of a proper definition of an

acknowledged hazard, was not consistent with the exercise of reasonable skill and care. The

experts were agreed that erodible rock was present in the HRT. The judge failed to give

adequate reasons about reasonable skill and care based upon his own rhetorical question:

“Why go to the bother (sic) of identifying such features and then decide to do nothing about

[154] The pursuers advanced indemnity as an alternative route (1A) to success. In terms of

clause 82, the defenders were required to replace and repair loss and damage. Clause 83

provided an indemnity in respect of the costs. The two routes ended up at the same point; a

damages claim for the refusal to carry out remedial work, and the indemnity. When there

was loss and damage attributable to a defect which existed at take over, that was a

contractor’s risk, whether or not, in terms of that clause, the contractor was obliged to

correct it. Clause 83 provided an indemnity, depending on whether the event was a

contractor’s or an employer’s risk. Option M did not convert what would otherwise be a

[155] The joint insurance only applied to contractor’s risks (clause 84.2). If the defenders

were correct, the use of reasonable care would not only result in an indemnity in their

favour, but the event would also fall outwith the scope of the insurance. If the pursuers

were correct, the provisions worked in a sensible way. First, the contractor would do the

work straight away. If it was a contractor’s risk, both parties would have access to the

insurance monies to fund the works. The insurance claim was expected to yield a maximum

[156] The commercial judge’s reasoning was short and omitted discussion of key legal and

factual disputes. The judge required to address the principle issues of law and fact (Dingley

v Chief Constable, Strathclyde Police 1998 SC 548 at 554 to 556; McLeod’s Legal Representative v

Highland Health Board 2016 SC 647 at paras 91-96). The parties had required to spend a

significant length of time going through the evidence at the hearing of the reclaiming

motion, which highlighted that the judge’s reasoning was deficient. The judge ought to

have recorded the principal evidence upon which he relied. An opinion should make clear

the basis upon which facts have been found. When a finding had been made on disputed

evidence, the opinion ought to explain the basis for that. The judge’s Opinion was stripped

back and bereft of the critical evidence. Only if adequate reasons had been given, could the

question, of whether the court of first instance had been “plainly wrong”, be answered. An

appellate court could interfere with factual conclusions if there has been a demonstrable

failure to consider relevant evidence (Henderson v Foxworth Investments 2014 SC (UKSC) 203

at para 67). The judge had failed to give adequate reasons for his central findings.

[157] The joint minute had agreed that certain documents were what they bore to be, but

that did not mean that their contents could be taken as true and accurate (McEwan v Lothian

Buses (supra) para 3). The commercial judge erred in proceeding on the assumption that the

REC sheets and the mapping were true and accurate. He wrongly took into account the

conclusions of two documents, which had not been spoken to by their authors viz.: the

minutes of the meeting noted by Mr Speirs; and the report by Mr Fawcett. The person to

whom this report had been put (the pursuers’ later project manager) had disagreed with it.

Dr Wilhelm’s evidence, that the mapping showed the criss-crossing of the shears and thus

the mechanism of the collapse, went without comment by the judge. Instead, he used the

[158] An expert could not give evidence on what specific terms marked on the REC sheets

meant, or what factors were taken into account in making the notations. The defenders had

been unable to elicit the evidence, which they required to do, because the pursuers had not

called the witnesses to speak to the documents. There were numerous examples of witness

testimony contradicting the documents. Where a whole raft of witnesses had, without

explanation, not been called, it was inappropriate to ask experts to give the primary

[159] The commercial judge’s conclusion about causation was irrelevant. There was no

requirement to prove that the exercise of reasonable care would have led to the installation

of a particular type of support, which would have prevented the collapse. If the collapse

had been due to a defect, which the defenders were obliged to repair, there was no need to

prove causation. If the collapse was a contractor’s risk, then the pursuers were entitled to

the indemnity under clause 83.1. The defenders were liable irrespective of causation. If the

[160] The commercial judge had erred in holding that the secondaries were not defects.

Donaldsons had drawn up a list of 114 areas where erosion had occurred. Jacobs had

recommended that there be extensive work done in these areas. The judge had excluded the

cost of this work, which was about £21 million, from the damages calculation.

[161] Caution was a factor which could be taken into account in determining what was

reasonable (see Governors of the Hospitals for Sick Children v McLaughlin & Harvey [1987] 19

Con LR 25 at 105). Whether the secondaries were a defect in their own right was

immaterial. Having had to carry out extensive repair work, the defenders had to be prudent

and eliminate the risk of erosion elsewhere. Dr Billig had conceded that 65% of the

secondaries required remedial attention. If this work required to be done, then the cause

constituted a defect. Dr Billig had referred to 74 of the 114 secondary features being those

which he would have expected to occur following a first dewatering. Although they posed

no immediate risk to stability, they ought to have been, and were, dealt with in order to

[162] The commercial judge had erred in not awarding the pursuers greater damages in

respect of the defenders’ breach of clause 82.1. If the defenders had been obliged to carry

out repairs at their own expense, the pursuers were entitled to the full cost of those repairs

as damages. The measure of damages was the difference between the costs incurred and the

sum which would have been paid to the defenders had they carried out the repairs. The

judge erred in holding that there was no evidence on this matter. The defenders’ evidence

was that the sum which they would have claimed was £28,729,000, plus a profit of

£3,102,470. The total was the amount that ought to have been deducted from the total paid

to BAM in computing the damages. It was accepted that the weakness in this argument was

that the evidence from Mr Hunter concerning the £31 million had been discredited.

[163] The commercial judge had erred by awarding the defenders 95% of their expenses.

There was no rational explanation for this and, in particular, for isolating the evidence about

the defenders’ lack of intention to repair the works as the sole matter justifying a discount.

The defenders had been unsuccessful on a number of other issues, including: (a) the

counterclaim; (b) the allegation that the pursuers were in breach by failing to give the

defenders an opportunity to carry out the repairs; (c) contributory negligence; (d) the

reasonableness of the remedial works; and (e) the quantification. About 40% of the proof

had been occupied by issues on which the defenders had been unsuccessful. But for these

issues having been raised, the proof could have been concluded in one court term. The

relevant principles were set out in William Nimmo & Co v Russell Construction Co (No. 2) 1997

SLT 122. The judge was “plainly wrong”. The pursuers had suggested an allocation of

50/50 on the basis that the defenders had succeeded in their defence on fairly narrow

[164] Accordingly, the reclaiming motion ought to be allowed and decree pronounced

either for £125,963,845.56, being the repair costs less amounts for inefficient working and the

repair of the culvert, or a sum which would be £22 million less in respect of the secondaries.

Paragraph 3 of the commercial judge’s interlocutor of 2 February 2017 should be recalled.

The court should sustain the pursuers’ second, third, fourth and fifth pleas-in-law, with the

[165] On the assumption that the event was a contractor’s risk and thus covered by

insurance, the authorities (CRS (supra)) established that, where joint insurance excluded

claims between the parties following an insured event, it did not exclude liability for breach

of a distinct obligation to repair the damage caused (see also the “Ocean Victory” (supra)).

The existence of an all risks policy was not an answer to a refusal by a contracting party to

carry out his obligations. The defenders’ breach in refusing to carry out repairs was not an

insured risk. The duty under the contract was to correct and not to compensate (CRS (supra)

at para 6). The balance between the parties in CRS could be found in the present contract.

The pursuers stood to lose £20 million per year as a result of the tunnel collapse, because the

contract generally excluded claims for consequential losses. The contractor took the risk of a

defect occurring, but his liability was limited to the low availability sum. The employer

would have no claim for compensation, but he could insist that the contractor carry out

[166] Joint insurance was irrelevant to a claim based on a breach of the obligation to

correct a defect. There was a distinction between a contractual scheme, which curtailed

recovery irrespective of the cause of the loss, and a contractual provision which backed up

other obligations enforceable against the other party by other means. The joint insurance in

the “Ocean Victory” had been comprehensive, exclusive and applied irrespective of the cause

of the loss (ibid para 114). That was not the position in this case. The parties had anticipated

a shortfall (clause 85.4) to be allocated to the party on risk. The defenders’ interpretation

[167] The contract clauses were of a different nature to those in the “Ocean Victory” (supra).

The remedies under the contract were not exclusive. The defenders’ approach attempted to

qualify the indemnity in clause 83, by reading into it that a contractor’s risk would not be

included because of the existence of joint insurance. The insurance provision was intended

to support obligations which remained enforceable. The insurance provision provided a

supplementary fund, akin to that in Surrey Heath BC v Lovell Construction (1990) 48 BLR 108

[168] The commercial judge had correctly construed clause 82.1. The defenders were not

entitled to refuse to carry out the remedial works until they had been paid, or payment was

agreed, if they considered that the collapse was prima facie an employer’s risk event. The

purpose of clause 82.1 was to secure the prompt repair of damage, not to delay the works

until any dispute were resolved (Thomas, Keating on NEC3 at para 9-019). The contractor

would bear the initial financial outlay. The adjudication clause (clause 92) provided that,

pending the resolution of the dispute, the parties were bound by the instructions given.

[169] The pursuers had not been in breach of their obligations in issuing an ultimatum to

the defenders, nor was there a break in the causal chain between the defenders’ breach and

the loss. It was only after the defenders had given the impression, that they did not intend

to carry out the works, that the pursuers began to explore the possibility of having someone

else carry it out. Had the defenders not breached their obligation, no ultimatum would have

[170] The commercial judge was correct to reject the counterclaim on the principle of

mutuality. Caution should be exercised when artificially dividing up obligations (Inveresk v

Tullis Russell 2010 SC (UKSC) 106, at para 43). Had the defenders carried out the recovery

project, the pursuers would have paid them, had the collapse been an employer’s risk event.

When the defenders refused to carry out the works, the pursuers had paid BAM. Since the

defenders were in breach of clause 82.1, they could not claim the costs of shadowing

somebody else who was performing what was their own obligation. This applied to both

loss of profits and the monitoring of the remedial works. For the litigation and adjudication

costs, it was not possible to predict what the dispute would have been between the parties,

[171] The causal chain between the defenders’ loss and the collapse of the HRT had been

broken by the pursuers instructing BAM and by the defenders’ decisions to shadow the

remedial works, to adjudicate twice, and to conduct this litigation. These were all voluntary

commercial decisions which were not the direct and natural consequence of the collapse.

The collapse required to be the effective cause (ENE Kos I v Petroleo Brasileiro (No. 2) [2012] 2

[172] The defenders had not proved quantum of their loss. They were not entitled to rely

on the evidence of Mr Hunter. The commercial judge found him unreliable. The defenders

had failed to prove that the £9 million claimed was reasonable or was causally related to

[173] There was an automatic right to be refunded an adjudication award, if the original

payer was successful in a subsequent litigation (Aspect Contracts v Higgins Construction

[2015] 1 WLR 2961, at para 23-24, 32). The commercial judge was correct in holding that,

having determined that no sums fell due to be paid under the counterclaim, the pursuers

[174] There was evidence of a period of delay due to the defenders’ refusal to carry out the

investigations. Had the defenders acted properly, the investigations would have been

finished by mid-December 2009, whereas they had not even commenced until then. The

defenders were not entitled to claim under the indemnity, as the principle of mutuality

meant that they could not recover what had been caused by their own breach.

[175] The commercial judge proceeded on the agreed basis that he would deal with

quantum on a headline basis. Thereafter the defenders had declined to participate in any

discussion. The parties had been asked to agree the level of costs, but agreement not been

possible. The parties’ experts had met to identify costs. Disputed costs had been set out in a

joint statement. The judge had deducted the sum of disputed costs (approx. £533,000) from

[176] The pursuers had proved the case to the requisite standard, taking into account that

the parties had engaged in a proof to identify the key issues of principle. The commercial

judge was entitled to take a top down approach to quantum, having rejected the evidence of

Mr Hunter, which proposed the opposite. The judge required to look at the reasonableness

of the repair work, which was what he did. Axa Insurance UK v Cunningham Lindsey United

[177] The DAT had not only been an acceleration measure. It had also brought benefits in

terms of ventilation, access and safety. There had been a duty on the pursuers to mitigate

their loss by repairing the tunnel as soon as possible, even if an early repair would cost more

(Governors of the Hospitals for Sick Children v McLaughlin & Harvey (supra). The issue did not

require to be determined, given that it was incorrect to describe the sole purpose of

constructing the DAT as mitigating loss of profit. The ability to clear the full 600-metre tail

of the collapse did accelerate the works, despite the fact that it did not shorten the time to

build the BPT by allowing drilling from both ends. The reasonableness of the works had to

be assessed by reference to the circumstances prevailing at the time and not with the benefit

of hindsight (Shepherd Homes v Encia Remediation (2007) 110 Con LR 90, at para 478).

[178] The commercial judge had erred: (1) in finding that the signal for needle valve

opening had not been transmitted to the control room in Perth; and (2) in addressing

contributory negligence, but not “fault” under clause 80.1. Whether “fault” was a higher or

lower standard than negligence was irrelevant. For there to be fault, the defenders required

to point to a contractual duty incumbent on the pursuers to complete the operational notes

(clause 83.2). Fault was different from “default” (cf City of Manchester v Fram Gerrard (1974)

6 BLR 70). A breach of duty involved some form of legal obligation (Norrie: “Fault” in Stair

[179] There was no duty on the pursuers to prepare operational notes and therefore no

fault in failing to do so. The responsibility for giving the pursuers the relevant information

lay with the defenders. They had provided 52 lever arch files; none of which had contained

any entry relating to needle valve opening. The purpose of the operational notes had been

to distil the key matters that the pursuers’ operators might need. The pursuers’ employee,

namely Claire McConnell, who had been asked to prepare the notes, would have had to

have shown more insight than the defenders in order to discern that the valve readings

ought to have been included and that the readings observed by the defenders’

subcontractors, Andritz, had required immediate action to shut down the plant.

[180] The defenders’ argument proceeded on the false assumption that, had they been

asked, Andritz would have told the pursuers something that would have caused them to

take action by 7 May 2009. During testing on 4 May 2009, Andritz had observed a needle

valve opening of 86% and had done nothing about it. The commercial judge found that the

pursuers should be excused for failing to notice that the needle valve opening reading was

86%, because of the chaotic conditions. The tunnel was not in a “steady state” of operation

during the testing. It was understandable for the pursuers to attribute rogue observations to

[181] There was no basis to find that, if the problem had been discovered by 8 May, the

cost of the recovery work would have been only £5 million. That figure was premised on

assumptions which were rejected by the commercial judge. It assumed that it would have

been possible to tunnel through, if the collapse had been smaller. Even Prof Broch and

Dr Palmström had accepted that a BPT had been reasonable. The possibility of tunnelling

through failed to take into account the dangers in doing so. It presupposed that the collapse

in the crown of the tunnel was only 8 to 10m, whereas the judge found that it was

[182] The commercial judge rejected Mr Hunter’s evidence, partly because he had

assumed too much certainty. Mr Hunter did not have the relevant expertise. The judge

rejected his methodology for calculating that the remedial works would have cost only

£7.5 million in August 2009. He had used the same methodology to arrive at the sum of

£5 million as an estimate for the works in May 2009, based on Dr Büchi’s best guess.

[183] The defenders maintained that there was no merit in the pursuers’ construction of

[184] First, the words required to be given their ordinary meaning. “Not liable for defects”

referred to the liability to correct defects under sections 4 and 8. Option M turned the

contract into one where there was no liability for design, if reasonable skill and care were

exercised. Section 4 made the fitness for purpose character of the original contract plain.

The onerous nature of that was accentuated by the addition of the 12 year defects liability

period; hence the need for, and the purpose of, Option M. Option M stated that the

contractor was not liable for certain defects; not that he was not liable for compensation due

to defects arising after the 12 year latent defects period. If Option M was a limitation arising

only after the end of the latent defects period, it was meaningless. There was no liability of

[185] Secondly (ground 4), the choice of what support to install had been part of the

design. It was a holistic exercise of judgment, taking account of the observed conditions. It

was part of the process set out in the contractor’s design and accepted by the project

manager. The judgments were evidenced by the REC sheets, which were referred to within

[186] Thirdly, on onus of proof (grounds 5 to 10), the pursuers’ general complaint, that the

defenders had not called the necessary witnesses, was bound to fail. It was only an

exceptional case that ought to be determined on the basis of onus. The commercial judge

had heard a wealth of evidence on which he was entitled to reach his findings in fact. At the

forefront was the finding that all the design decisions taken in the tunnel had been approved

by Jacobs. This finding impacted on many different aspects of the case, including the

pursuers’ ability to establish a defect in design, onus and reasonable skill and care. The

[187] The contractual analysis should not be based simply on the documents, but on the

evidence about what everyone had done. The accepted design referred to a process. The

meaning of design was informed by the evidence which explained the working of the

process at each stage. The commercial judge had correctly understood the contractual and

[188] The responsibility for the design rested with the defenders (clause 14). An essential

element of design was choice (Hudson: Building and Engineering Contracts (13th ed), at

para 378). The project manager had accepted not merely a document, but a process. If

everything that happened in the tunnel were left out of the equation, and the design were

simply what was contained in the documents, all that would exist would be the drawing

[189] This was a re-measurement contract. If there were an increase in the class of

support, the contractor would be paid for that. If the re-measurement was less than the total

estimate of £129m, less would be payable. There had been no incentive on the defenders to

provide less support. There was an economic benefit to the pursuers in decreasing it.

[190] The project manager had certified the tunnel as complete for safe operation in

December 2008. The design was accepted by this process. The contractor had done all that

the works information had required by the completion date. Defects under section 4 were

shortcomings; loss and damage under section 8 in contrast were fortuitous. There were

[191] In describing both the obligation to correct defects and the obligation to pay for

correction, the word “liable” was used. It was the same word as that used in Option M. In

the case of latent defects, the contractor submitted proposals to remedy the defect. If the

contractor failed to remedy the work, the contractor was “liable” for the costs of employing

another. On the pursuers’ version, there was a period during which both regimes would

operate. The pursuers’ suggestion that Option M limited liability beyond the latent defects

period was not well-founded, since there was no liability at all at that stage.

[192] If an upgrade were required, that amounted to a defect; a defect in design being

something that was not in accordance with what ought to have been proposed. If, following

dewatering, a defect had been detected, the obligation to correct it would continue, subject

to Option M. If what had been installed was class IIA, and it should have been class III, that

[193] Section 8 dealt with risks and insurance. This was a different regime from defect

correction. Loss and damage after take over was an employer’s risk, unless it was due to a

defect which existed at take over. There was no such defect and, in any event, Option M

operated to exclude liability. The pursuers were treating them as alternative routes to arrive

at the same conclusion, but they were not the same. The commercial judge would have

[194] The heading to Option M in the NEC contract was “Limitation of the contractor’s

liability for his design to reasonable skill and care”. There was no reference to defects. The

standard form was intended to be suitable for parties entering into a design and build

contract. It was unsurprising that Option M was the clause which achieved its object as part

of a standard menu. It was not possible to get anything simpler than a statement that the

contractor was not liable for defects. Option M took priority over section 4 because the

standard form already contained section 4. The drafters had provided an easy way of

converting the contract for the purpose of design, rather than construction; the latter

[195] The alternative construction denuded Option M of any real purpose in a design

situation, unless a contractor was prepared to accept that the contract remained one of

fitness for purpose. The effect of Option M was that it cut through section 4. The exception

in clause 80.1 tied sections 4 and 8 together. Loss and damage to the works taken over by

the employer would be an employer’s risk, except where it was due to a defect which

existed at take over. The contractor was not liable for loss and damage due to defects which

[196] The onus of proving that a defect existed remained with the pursuers. The fact that a

design life had been stipulated did not mean that the life of the works was guaranteed (MT

Højgaard v E.ON Climate & Renewables UK Robin Rigg East [2017] Bus LR 1610, paras 6, 27-

32, 44-45). It would have to be demonstrated that what had been done in design terms could

not, or was not going to, produce a scheme which would function for that period. That was

not the case here. In any event, the 75 year period had been varied in terms of Schedule

Part 7, para 6, which applied it only to concrete structures. The tunnel was not such a

structure. There was no design life for the civil works in general. The commercial judge had

[197] The pursuers’ attempt to classify the pre-watering up inspections as an “event” was

contrived. The inspections were part of the works. They were not contractually significant,

[198] Even if it was accepted that the design was not a process but the content of

documents, the latter said only that support was to be installed to the extent required to

meet the conditions encountered. The judge found that the defenders had installed that

support. There was no disconformity between such implementation as the design required

and the implementation which was carried out. If what had been decided in the tunnel

were not part of the design in contractual terms, it was in practical terms. Option M limited

liability for all design, not simply design which had been accepted in terms of the contract.

[199] Everyone knew that they were contracting for an unlined tunnel using the

observational method and that a risk remained. The pursuers had tried to make the contract

work by engaging with one of the world’s top contractors, namely the defenders, and one of

the world’s top consultants, that is Jacobs. The risk remained but it could not fall upon a

party who could only do what was agreed in light of what could be observed.

[200] Once evidence had been heard, questions of onus rarely arose (Salt International v

Scottish Ministers 2016 SLT 82, paras [45]-[49]). Cases which fell to be decided upon onus

had to be exceptional (“The Popi M” (supra); Dingley v Chief Constable, Strathclyde Police

[201] The pursuers’ complaint ignored the significant body of evidence to which the

commercial judge had regard. The judge gave (para [174]-[177]) four reasons, which formed

the basis for his conclusion that reasonable skill and care had been exercised. None of the

experts, with the exception of Prof Sloan, had been prepared to impugn the decisions taken

in the tunnel. The judge had seen the primary evidence of mapping and the REC sheets. He

had heard the testimony of what others had made of this. He rejected (para [187]) the

pursuers’ contention that the defenders had not exercised reasonable skill and care on the

basis that it had been founded upon hindsight. All of the points which had been raised by

the pursuers in their submission had been made to the judge. Hills v Niksun (supra) was

distinguishable. When all of the evidence, notably that of the experts, based on all of the

detailed primary evidence, was looked at, it could not be said that the judge did not have

evidence that entitled him to reach the conclusion that reasonable skill and care had been

[202] The commercial judge had not just been faced only with expert evidence. He had

heard from the members of the TBM crew who had spoken to the process carried out in the

tunnel, the work of Mr Taylor and the role of Jacobs. They had described the care taken by

Mr Taylor and what could be seen in the tunnel. Mr Sandilands spoke about the role which

both Jacobs and Pöyry had played. The judge had heard from others who had inspected the

tunnel, notably Prof Broch. Prof Müller, the defenders’ design co-ordinator, spoke to the

RSM, the REC sheets and the risks which had to be taken into account.

[203] In their pre-proof preparations, the pursuers had asked the defenders if there was

any objection to them speaking with Mr Taylor. There had been none. One condition had

been that the pursuers provide the defenders with Mr Taylor’s details, since they had lost

[204] There was no merit in the individual points taken by the pursuers. The pursuers

founded on the evidence of Dr Wilhelm, but he was giving evidence with the benefit of

hindsight. He was explaining that, although the shear zones on the mapping would not

explain the collapse, they must have been interconnected above the HRT. He did not say

that the collapse could have been anticipated. The commercial judge listened to the

pursuers’ submission on precisely this point and made positive findings that nothing could

[205] In his email of 28 August 2007, Mr Taylor had recognised that, during the time when

the TBM crew were in the CFZ, it would be prudent to maintain a minimum class of

support, even although the rock conditions, superficially, may have looked like class I.

There was nothing in this that involved missing something indicative of a potential collapse.

The commercial judge rejected the idea that Mr Taylor’s reference to specific hazards during

excavation meant that the REC sheets had failed to address long term risk. The judge was

entitled to take into account the fact that no problems had been encountered when the TBM

[206] The REC sheet for chainages 2084 to 2082 had a majority of ticks that were either

class II or III; yet the overall class was I. This area had been only 2 metres wide. There was

[207] Mr Speirs’ minute of the meeting on 9 September 2007 supported the commercial

judge’s conclusion that there had been nothing to be seen in the CFZ. The document, which

was agreed as being what it bore to be, had been referred to by Mr Sandilands. It was there

[208] Simply because kakirite was recorded on mapping, and ought to have been

mentioned on the REC sheets, did not mean that class IV had been required. Merely

because the mapping and the relative REC sheet were not identical, did not mean that they

were inconsistent. The alternative explanation was that, as part of the holistic exercise, any

kakirite had been so minimal that it did not create an issue about the support level.

Everyone would have seen the reference to kakirite and none had suggested that there

ought to have been class IV support. The REC sheets and the mapping were agreed to be

what they bore to be. That did not mean that they were true and accurate. It was up to the

commercial judge to determine that. He was not just given them and asked to reach a

conclusion based on what he could see. He had the testimony of those who had described

them. Prof Broch had investigated the locations described by Pöyry. He had found no

major weakness zones and considered that the support measures were more than good

[209] The reference to chainages 2117 to 2101 was another isolated attempt to undermine

the exercise of judgment at one location. The commercial judge held that the Pöyry findings

had been minor. Those in the HRT had inspected not just the areas focused by Pöyry, but

also the whole of the tunnel. The suggestion, that if Pöyry had inspected the whole tunnel

the result would have been different, was destroyed by the finding that there was nothing to

[210] In relation to the dip of <500, even Prof Sloan had not categorised this area as class IV.

Dr Büchi had explained that, although there might have been some areas which pointed

towards class III, that did not mean that the overall rock mass should be classified as such.

The classification was a summary of all aspects. Dr Büchi did not consider that the mapping

and the REC sheet meant that the area should be class IV overall. The commercial judge’s

use of the rhetorical question: “Why go to the trouble of identifying such features and then

decide to do nothing about them?” was based on the evidence of Dr Büchi and Prof Sloan’s

[211] The commercial judge found that Table 11 did not place a mandatory obligation on

the defenders to shotcrete all erodible rock. The pursuers’ contention, that class III or IV

support was required whenever the defenders discovered erodible rock, was indicative of a

painting by numbers approach. If the pursuers were correct, determination of overall class

would have been reduced to a mechanistic process, dependent on the worst feature ticked.

The process was, on the contrary, a holistic exercise of engineering judgment. There was

positive evidence in favour of the conclusion that the risks stated on Table 11 had been taken

into account. Just because the REC sheets contained no express reference to Table 11, did

not mean that the issues raised in it had not formed part of the engineering judgment.

[212] The evidence did not support the pursuers’ contention that there was a two-stage

system involving, first, excavation and, secondly, operation. The commercial judge referred

to the pursuers’ contention that the REC sheets had only addressed hazards during

excavation as “startling”. He posed the rhetorical question: “why would the defenders

construct a tunnel with no thought to its long term stability?” In the email, Mr Taylor had

simply been explaining the dangers during excavation. The judge was entitled to find that

the proposition was startling because of two tracts of evidence. First, Prof Müller had said

that everything had been covered in the REC sheets. Everybody had known that they were

engaged in building a hydro-electric tunnel. Everyone had known whether they were in the

HRT or the TRT. It was obvious that, when building a pressurised tunnel, account had to be

taken of that element. Secondly, Dr Wilhelm had said that working safety was an issue

during construction, but rock classification had to take into account overall stability during

operation. The judge had heard that the defenders, Jacobs and Pöyry, who had inspected

the tunnel in January and February 2008, were actively looking for problems. Prof Broch

had inspected the tunnel. The proposition, that all of these people were only considering

health and safety during excavation, was absurd. Dr Büchi spoke to the REC sheets dealing

with operational matters, including erosion. Even if Prof Broch had supported this theory at

the Adjudication, he was only one witness and he had not been involved until after the

February 2008 inspection. The Pöyry guideline, which required observation of the

behaviour of the rock behind the face, should be regarded not as a two-stage process, but as

part of a ongoing process of checking and revising. The judge took it to be consistent with

the defenders’ position that it included taking into account operational risks. The extremity

of the pursuers’ position was that, in building a hydro-electric tunnel, the defenders, Jacobs,

Pöyry and Prof Broch had not considered the effects of flowing water under pressure. The

lack of a response to the query from Jacobs about the definition of erosion was only one of

many factors involved in the judge’s decision. There was no evidence to suggest that the

parties had had any difficulty in understanding the meaning of erodible rock.

[213] The commercial judge made very detailed findings about the 2008 inspections.

Prof Broch had observed the tunnel only 13 days after the Pöyry inspection. He had looked

at the locations described by Pöyry and had walked the parts of the tunnel in between.

There had been joint inspections thereafter, when Mr Speirs had expressed a concern about

the lower side wall, but had been reassured that, if erosion occurred, it would not affect the

stability of the tunnel. The argument, that Pöyry would have noticed something in the

collapsed zone and recommended installing something major, ignored the judge’s findings

in fact that: (1) there was no problem to be seen in the tunnel; and (2) the issues which gave

rise to Pöyry’s recommendations had been minor and all had been resolved prior to

watering up. The Pöyry report had given rise to defects notice D033, which had been

[214] The pre-watering up inspection had been to clean out the tunnel to make sure there

was nothing in it which could flow down and damage the turbine. It was not an inspection

of the geology or support installed. The erosion noticed in October 2008 was not the same as

that which had been noticed earlier, given the closing off of defects notice D033. What was

noted was what had been expected by way of rockfall. This was consistent with defects

notice D041, which recorded a significant quantity of gravel, sand and silt in the rock trap. It

noted the presence of significant voids below the invert, where weak and/or sheared rock

had been eroded by the water flushing exercise. Defects notice D044 had recognised what

required to be done and that it was appropriate to hold off remedying the defects until the

first dewatering. The witnesses had stated that there were no concerns about the watering

up on 27 November 2008. The points raised by the pursuers were all miniscule matters that

had to be looked at in the context of an 84 day proof. The commercial judge heard them and

made nothing of them. Looking at them on the printed page, it was readily understandable

why he reached that view. If the judge did not give the pursuers’ points greater prominence

in dismissing them, that reflected their inherent weakness. It cannot be said that he was

[216] Although he had summarised his reasoning adequately, the commercial judge must

have had regard to the “penumbra of imprecision” (Biogen v Medeva [1997] RPC 1 at 45). An

appellate court was bound to assume that a judge has taken into account the whole of the

evidence (Henderson v Foxworth Investments 2014 SC (UKSC) 203 at para 48; Thomas v Thomas

1947 SC (HL) 45 at 61; and Housen v Nikolaisen [2002] 2 SCR 235 at para 72). It was entirely

appropriate for the judge to summarise and marshal points at an appropriate level of

generality, with a view to focusing the issues and not getting lost in the detail (McLeod’s

Legal Representative v Highland Health Board (supra), para [93] ). Read as a whole, the judge’s

Opinion was far from being an oracular pronouncement (McLeod’s Legal Representative

(supra) at para [91]; citing Dingley v Chief Constable Strathclyde Police (supra) at 555).

[217] The commercial judge’s findings on the mapping and REC sheets did not proceed on

an assumption that they were true and accurate, but on the basis of the evidence. The TBM

crew had referred to the sheets having been completed by Mr Taylor in consultation with

Jacobs. Mr Sandilands had expressed his view on the importance of the REC sheets and

Jacobs’ signature binding the pursuers. The experts had spoken to the high standard of

[218] The commercial judge had been entitled to take into account the terms of the minute

noted by Mr Speirs and the report from Mr Fawcett. The former had been referred to by

Mr Sandilands and the latter by his successor, Mr Brand. Mr Brand had referred to

[219] It was a curiosity that causation was a miscellaneous ground of appeal, when it was

so crucial. The experts were generally agreed that only class IV support would have

prevented the collapse, yet no one had suggested the use of class IV at any point in the

tunnel. The only class III that was ever installed was at chainage 2120, where increased

shotcrete had been installed. The pursuers had maintained that at least class III should have

been installed. The commercial judge accepted the preponderance of the evidence, notably

that given during the hot tubbing, which favoured the conclusion that only class IV would

have prevented the collapse. The judge accepted the evidence of four out of the six experts.

That was not something which could be regarded as “plainly wrong”.

[220] It was not enough simply to say that the defect consisted of insufficient support. The

pursuers had to point to the defect by reference to the observational method. The pursuers

required to prove that the loss and damage had been due to a defect which existed at take

over. There had to be a causative link between the defect and the loss and damage. This

issue was no different in connection with the correction of defects under section 4, where the

pursuers were seeking the cost of having the defect remedied. The pursuers had to establish

that these costs were those for remedying the defect and not for remedying a lack of

support. Demonstrating that the collapse would have been prevented had the defenders

installed the support necessary to avoid that collapse, did not identify what, or where, the

defect was. The pursuers had to be able to say that the defenders had to incorporate class IV

support at the point of collapse. If the obligation was only to shotcrete erodible rock, the

[221] The commercial judge found that the secondaries did not constitute defects.

Attempts had been made to locate the secondary defects, but Prof Broch, who had inspected

the HRT in January 2011, had concluded that they were all either rock features which could

be safely left unsupported or minor rock fallouts which might be expected on a first

dewatering. Dr Billig’s analysis was that none posed a risk to stability. Although at one

point Dr Billig had said that he had not thought about the operation of the tunnel, it was

clear that he was considering its load carrying capacity. He had discussed erosion, which

[222] The commercial judge had recorded that the point about the secondaries had not

been pressed by the pursuers, although it was now. On 13 April 2010, the defenders,

Donaldsons, Jacobs and BAM, had all gone into the tunnel to find the faults recorded in the

defects notice D058. They had failed to do so. Even after the problem about the revised

chainages had been resolved, Mr Becker, who had been one of the defenders’ staff, had

testified that the features did not look like defects. The evidence was all consistent with

[223] The contention that it was reasonable for the pursuers to repair secondary defects,

given the nature of the collapse, was a view which might be taken from an economic

standpoint but it was misconceived from a legal perspective (Governors of the Hospitals for

Sick Children v McLaughlin & Harvey (supra)). The defenders were not liable for the

[224] The argument that, because the defenders had said that they could repair the tunnel

for £30 million, they should get the difference between that and the actual cost of repair,

ignored the fact that the £30 million figure had been based on the discredited evidence of

[225] The commercial judge had erred in finding that the defenders had been in breach of

clause 82.1. He ought to have found that the pursuers had placed unlawful pre-conditions

on the defenders. The pursuers had insisted upon a concrete invert, the DAT and BPT.

They had wanted this work carried out as quickly as possible in order to mitigate their loss

of profit. They were attempting to impose an ultimatum that the design for the remedial

works had to incorporate three specific aspects. The pursuers would not have allowed the

[226] The commercial judge’s decision constituted a reasonable exercise of discretion.

Success was a relative term and great latitude was allowed in defining it (McLaren: Expenses,

p 22; Gye & Co v Hallam (1832) 10 S 512). In general, awards would not normally be made

for different aspects of a proof. McLaren post-dated the introduction of the tender system

[227] The starting point was the “vindication of rights”. The cost of litigation should fall

on the person who has caused it (Shepherd v Elliott (1896) 23 R 695, at 696). Success was not

capable of determination according to rigid rules. It was ultimately a matter of discretion

(Howitt v W Alexander & Sons 1948 SC 154, at 157). Where there was divided success, there

could be many permutations (ibid at 158; McFadyen ed.: Court of Session Practice at para L103).

If there were distinct, severable branches and the parties had been successful in different

[228] The pursuers’ award for low availability damages, which took up no time in

evidence and minimal time in submissions, was a very small percentage of the sum sued for.

It involved a minor skirmish in the context of the case as a whole. It was a Pyrrhic victory in

circumstances where the real success was on the part of the defenders (ibid; D Macdonald &

Bros v Cosmos Decorators 1969 SLT (Sh Ct) 9 at 10). If the sums which the pursuers had been

awarded had been offered pre-proof, the litigation would still have gone ahead. The fact

that the defenders could have tendered was a factor to be taken into account, not in

determining the question of success, but in the modification of any award (Sidlaw Industries

[229] The deduction of 5% from the defenders’ expenses was a proper exercise of his

discretion. The 5% discount was not de minimis. It represented a very significant amount of

[230] The commercial judge had erred in holding that clause 84 did not preclude a claim

by the pursuers against the defenders for loss and damage due to a contractor’s risk. By

entering into clause 84, the parties had agreed that they would look only to the insurance

monies in respect of risks falling within the policy. There was a longstanding rule that an

insurer could not, by way of subrogation, sue one co-insured in the name of another

(Commonwealth Construction Co v Imperial Oil (1977) 69 DLR (3d) 558 at 561; Petrofina (UK) v

Magnaload [1984] QB 127; Stone Vickers v Appledore Ferguson Shipbuilders [1991] 2 Lloyd’s Rep

288; and National Oilwell UK v Davy Offshore [1993] 2 Lloyd’s Rep 582). Since CRS (supra), it

had been clear that the rule was founded on contract. A term fell to be implied into the main

contract (Hopewell Project Management v Ewbank Preece [1998] 1 Lloyd’s Rep 448; CRS at

paras 48-50; Scottish & Newcastle v GD Construction (St Albans) [2003] Lloyd’s Rep 809; The

Board of Trustees at the Tate Gallery v Duffy Construction [2007] Lloyd’s Rep 758).

[231] An implied term could not withstand express language to the contrary (Tyco Fire &

Integrated Solutions (UK) v Rolls-Royce Motor Cars [2008] Lloyd’s Rep 617 at 76). If an

underlying contract envisaged continuing liability on the part of one co-insured to another,

even within the sphere of the cover provided by the joint policy, subrogated claims would

not necessarily be precluded (ibid at 77). In the “Ocean Victory” (supra) the Court of Appeal

and the majority of the United Kingdom Supreme Court concluded that the principle was

best viewed as resting on the natural interpretation of, or implication from, the contractual

arrangements giving rise to the co-insurance (ibid at para 114). The commercial judge erred

by having regard to, and giving weight to, contractual terms which he considered to point

[232] Even the minority in the “Ocean Victory” (supra) had acknowledged that, where it

was agreed that insurance would inure to the benefit of both parties, they could not claim

against each other in respect of an insured loss. There was a well-established rule to that

effect. The distinction between the majority and the minority was whether the joint

insurance meant that there was no liability at all between co-insured, or whether the liability

was satisfied by the joint insurance provision. That was not in issue in this case.

[233] In a situation where a party had failed to take out insurance, as provided for by the

joint insurance clause, the risk would fall where it fell. The possibility of an insolvent

insurer was remote and could not be a factor within the contemplation of the parties. The

parties had not anticipated a shortfall, as they had agreed that the insurance was to be for

replacement cost. If a party failed to take out insurance, the effect of the joint insurance

provision in the contract was the same (Scottish & Newcastle v GD Construction (supra), at

para 28). The rest of the contractual provisions could be read consistently with the joint

insurance provision. The commercial judge erred in concluding that certain terms were

[234] If the pursuers’ submission was correct, it would render clause 83 redundant. Under

clause 84, the insurance was taken out for a defined period. In contrast, the indemnity

under clause 83 was wide ranging, and included employer’s risks. It would allow for the

recovery of potentially substantial deductibles, which could arise. The inclusion of

clause 85.2 had been a “belt and braces” approach; to the make it clear that there was no

possibility of recovery against directors or employees either. There was no relevance to the

lack of any express waiver, and the commercial judge erred in taking that into account. The

judge’s decision deprived the parties of the protection of the implied term on the basis of

minor inconsistencies with different contract terms. Those terms could have been

reasonably interpreted as applying to different contractual obligations (Reed: Construction

[235] The effect of clause 84 should be viewed in the context of a coherent regime in

relation to liabilities. Where the contractor failed to correct defects, the employer was

entitled to recover the cost. The position was different in relation to loss and damage. The

responsibility for them depended upon whether the risk was that of the employer or the

contractor. The distinction between defects and loss and damage had a commercial

rationale. The contractor was responsible for carrying out the works so that they were free

from defects. He was therefore, subject to the provisions of Option M, responsible for

repairing them at his own cost. The position was different in relation to loss and damage,

[236] The commercial judge erred in holding that the defenders had been in breach of

clause 82.1 and prevented from counterclaiming successfully. The purpose of clause 82.1

was to prevent wrangling about payment until any repair had been completed. In order to

oblige the contractor to carry out the repairs, the employer required to present a prima facie

argument that the event had been a contractor’s risk. At the stage at which the defenders

had been called upon to carry out the repairs, the pursuers had not been able to present such

a prima facie position. The defenders had been entitled to rely upon the indemnity and insist

on payment. Clause 82.1 had to be read with clauses 60 and 80. The defenders’ position, as

set out in the letter of 25 September 2009, had been that they were ready to mobilise to carry

out the repairs in terms of clause 82.1. In terms of clauses 61.3 and 80.1, the cost of repair fell

[237] There were two elements to the counterclaim. First, there was indemnification under

clause 83.1 for costs which were due to an employer’s risk. These included the immediate

investigation of the collapse, the monitoring of the remedial works carried out by the

pursuers, the costs of the Adjudications, the cost of litigation and the management time

expended on dispute resolution. Secondly, there was the defenders’ entitlement to carry out

the repairs themselves. The commercial judge’s finding that the defenders’ costs had not

been incurred due to the collapse was wrong. The test was whether the collapse had been

an effective cause of the costs (ENE Kos I v Petroleo Brasileiro (No. 2) (supra), at para 12). It

had to be an effective cause, but it did not need to be the only cause. The legal costs and

management time were caused by the collapse. The Adjudications were a direct

consequence. The pursuers had raised this litigation in order to displace the results of the

Adjudication. The indemnity provided for the recovery of the costs of this litigation;

recognising that any modification by the court would break the causative link.

[238] The defenders’ obligation under clause 82.1 was not reciprocal to the pursuers’

obligation to indemnify. The costs to be indemnified, such as those of the Adjudications and

the litigation, bore no relation to clause 82.1. The defenders’ costs in successfully vindicating

their position were not dependent on clause 82.1. The authorities on mutuality were well-

known (Bank of East Asia v Scottish Enterprise 1997 SLT 1213, at 1216, citing Turnbull v McLean

(1874) 1 R 730, at 738). Each obligation by one party was not necessarily the counterpart of

every obligation by the other (Macari v Celtic Football Club 1999 SC 628, at 640). A party in

breach was not entitled to insist on the other party performing his counterpart obligations,

but caution should be taken to avoid splitting the contract artificially (Inveresk v Tullis Russell

[239] The commercial judge erred in obliging the defenders to repay the sums under the

second Adjudication. The sums fell entirely within the indemnity. The second adjudicator

had effectively permitted the defenders to enforce the indemnity under clause 83 on the

basis that these were costs which were due to an employer’s risk event. The pursuers had

argued that the defenders required to repay the sums as they had not effectively been

caused by the collapse. The argument had now evolved into an adoption of the judge’s

reasoning that the defenders were not entitled to payment because of their breach of

clause 82.1. The defenders had been entitled to retain the payment by the pursuers under

the adjudicator’s award as they had not breached clause 82.1. If there had been a breach of

clause 82.1, the payment fell into the category of costs which the defenders were entitled to

[240] The commercial judge erred in awarding the pursuers low availability damages of

£1 million under the terms of the bespoke clause 48.2. Clause 80 (Z11) provided that there

was no liability for indirect consequential loss, ie loss of profit. Clause 48.2 was an exception

to that general position, but it required to be read in conjunction with the indemnity in

clause 83.1. If the reason for the plant operating below the low availability standard had

been the collapse, and the collapse was an employer’s risk, the defenders were strictly liable

under clause 48.2, but were then entitled to the indemnity under clause 83.

[241] The pursuers’ claim had not been based on the defenders’ breach of clause 82.1, but

on the fact that the plant had been unavailable for many months. The commercial judge

treated it as a free standing claim, which did not depend upon any defect or breach of

contract. By the time of the By Order, the argument had shifted to focus on the low

availability damages as a consequence of the breach of clause 82.1. The judge awarded

damages because of the defenders’ refusal to return and carry out the recovery project in

breach of clause 82.1. If there had been no breach of clause 82.1, the entitlement to the low

[242] Even if there had been a breach of clause 82.1, there was no evidence to substantiate

the low availability damages. The pursuers would have required to plead and prove that a

part of the two year delay had been due to a breach of clause 82.1. The pursuers’

retrospective attempt to tie the damages to a breach of clause 82.1, as distinct from 48, did

[243] The commercial judge erred in concluding that the hypothetical sum, to which the

pursuers would have been entitled, amounted to in excess of £107 million. All the pursuers

had done was lead evidence of the money which had been expended on the remedial works.

They had sought to attribute that sum (£129 million) to various heads of claim and to

establish the reasonableness of the works carried out. The pursuers had not led evidence of

a causal link between the collapse and the sums which had been expended. The only

evidence had been that the remedial works had been reasonable, that a total of £129 million

had been spent on the works, and that certain deductions had been made by the judge on

the basis that the defenders could not possibly have been liable for those.

[244] The pursuers’ head of finance had given evidence about the pursuers’ accounting

systems. He had not carried out any exercise to ascertain which costs had been a

consequence of the collapse and which were not. The pursuers’ cost expert had provided a

detailed assessment of the costs of the recovery project. He had expressly said that he had

not looked at causation. The pursuers’ witnesses on monitoring had simply checked to see

whether the projected costs had been accurately forecast. The pursuers had contracted for

the remedial works on a costs plus a percentage basis; ie Option E. It did not follow that

each and every cost had been connected to the works carried out. The commercial judge

had erred in adopting a “top down” approach by starting with the sum claimed by the

pursuers and deducting sums which could never be attributed to the defenders (AXA

Insurance UK v Cunningham Lindsay United Kingdom (supra) at 276-287).

[245] The commercial judge erred in allowing the costs of the DAT to be recovered. The

purpose of the DAT was to allow tunnelling from both ends, to speed up the works, and to

reduce the time that the plant was out of operation. Recovery of the costs of the DAT would

cut across clause 48.1, which provided the only remedy for loss of profits. Clause 80 (Z11)

prevented any claim for consequential loss. The pursuers were entitled to take a commercial

decision to try to get the plant up and running as quickly as possible, but the terms of the

contract had to be applied in determining whether they were entitled to recover that from

[246] The commercial judge erred in rejecting the argument that the pursuers’ operation of

the plant had been an employer’s risk event. The pursuers ought to have noticed that the

needle valves had been opening beyond their normal parameters. The judge erred in fact in

finding that the needle readings had not been relayed to Perth. The signals had been

relayed, although they did not produce an audible alarm. The judge did not consider what

the consequences, of the needle valve signal being available in Perth, had been.

[247] The pursuers had been provided with a large number of maintenance and operating

files by the defenders. To make the information manageable, the pursuers had produced

operational notes. The notes were incomplete in relation to needle valve openings. Claire

McConnell, who had been responsible for preparing them, had accepted that they ought to

have said something about them. If information on the normal parameters had been in the

notes, the engineers would have noticed the developing trend. Acting with reasonable

diligence, they would have become aware of the collapse by April or May 2009. The extent

of the collapse would have been far less, and the cost of remedying it would have been

[248] Clause 83.2 provided for a proportionate reduction if events were contributed to by

the party seeking the indemnity. The commercial judge had characterised this as

contributory negligence, but it was really an issue of causation. If part of the pursuers’ loss

had not been related to the contractor’s risk event, but to their operation of the plant, there

ought to have been a proportionate reduction. Clause 80.1 provided that, included within

employer’s risks, were costs due to his negligence or fault. The failure to update the

operational notes had amounted to negligence, failing which, at least fault. Fault involved a

lesser degree of culpability than negligence (City of Manchester v Fram Gerrard (supra) at 90;

[249] It was not possible to give a precise quantification of the level of collapse in April or

May 2009. Dr Büchi had said that the extent of the collapse had increased each time the

plant had been operated. He had tried to estimate the total rockfall in April or May. In the

face of a lesser collapse, the appropriate remedy would have been to tunnel through at a cost

of just over £5 million. The pursuers had led no evidence to challenge either the theory

about the collapse being a lesser one, or the ability to tunnel through or the cost of that.

[250] There can be little doubt that the collapse of the HRT resulted in loss of, or damage

to, the works. The first question of contractual interpretation is whether, irrespective of

whether that loss or damage was caused by a defect, the defenders were obliged to replace

the loss, or to repair that damage, under the terms of clause 82.1, given that it was caused

before the issue of the defects certificate, due 2 years after take over. It is clear that the

defenders were indeed obliged to do this. Their refusal to do so was a breach of contract

[251] The second question involves identifying who was responsible for paying for any

replacement or repair to the loss and damage. That turns on where the risk lay. It would

rest with the pursuers unless the loss and damage was due to a defect which existed at take

over in terms of clause 80.1. The cost of any replacement or repair incurred by the defenders

would, other than in the event of such a defect, be subject to indemnification by the pursuers

[252] The third question is whether the collapse of the tunnel occurred because of a defect

which existed at take over; that is to say that either a part of the works was not in accordance

with the works information or a part of the works designed by the defenders was not in

accordance with the contractor’s design as accepted by the project manager (clause 11.2(15)).

[253] On the first element, the works information contains a general description of the

works (Sch Pt 3, sec 1, para 2.1). It was to be a hydro scheme providing reliable service

without the requirement of major refurbishment or significant capital spending within the

design life of the scheme. The design life for the HRT, which was a component of the civil

works, was 75 years (ibid para 6.3.2). This was not qualified by paragraph 6 of Schedule

Part 7, which did limit the life of some components but not the general requirement.

[254] The general requirement involved the creation of a tunnel with a 75 year design life.

That requirement did not guarantee such longevity. It meant merely that the design was

one which, if properly implemented, ought to have produced an HRT which lasted that long

(MT Højgaard v E.ON Climate & Renewables UK Robin Rigg East [2017] Bus LR 1610, Lord

Neuberger at paras 30 and 32). If what was constructed was not capable of lasting for the

design period, it would not be in accordance with the works information. If that were so,

there would be a defect in terms of the contact (clause 11.2(15)) and a failure to provide the

[255] MT Højgaard (supra) involved a requirement that wind turbines have a design life of

20 years. There was a two year period in which to give notice of a defect. It was said, no

doubt correctly, that this enabled the employers to use that time to inspect the turbines and

to detect any such defect. That is not dissimilar to the contract between the parties.

However, the defect in MT Højgaard was a miscalculation of the capacity of connectors in the

original design. The design as built could not have lasted for the stipulated life span. That

is different from a situation in which it is accepted, apparently by all, that the defenders’ (or

rather Pöyry’s) design for an unlined tunnel (ie the HRT RSM and Drawing D201, infra)

ought to have lasted for the 75 year period, at least in the ground conditions then

understood to exist and as set out in the relative report (Sch Pt 3; Appdx 12 supra). Since

what is under consideration is design life, it is only if an inconsistency between the design

(as accepted) and the potential 75 year life can be found that a defect can be said to have

[256] The commercial judge did not directly address the question of whether there was a

defect of this nature. His findings in fact are indicative of a fault existing at take over in that,

as the judge held (para [147]), the support provided was “not enough”; “poor rock

conditions coincided with insufficient shotcrete and rockbolts”. Given the actual state of the

rock, the HRT could not have survived for 75 years, or even 1 year, in the conditions which

in fact existed. The question then is whether the HRT can be said to be disconform to the

works information (ie the design life) by virtue of its (very) early collapse. The evidence,

beyond the occurrence of the collapse, does not support a contention that what was

designed was not capable of lasting the required 75 years. On the contrary, once the parties

had agreed upon an unlined tunnel, the general consensus of the experts and others was

[257] In terms of the works information, the detailed design (RSM and Drawing D201, see

infra) was to be “developed through construction detailing and method statements” (Sch P 3,

sec 1, para 6.2.8.4). The contractor was to map the HRT to enable the rock mass classification

system to operate (ibid, section 2, para 2.3.1). This was expanded with the terms of the Pöyry

design statement (Sch Pt 7; Appdx 6) which used this system. It was Pöyry who had said

that, with the use of the TBM, the HRT was feasible, with 60% unlined, without

compromising the design life (ibid para 2.4). What was to be constructed, using this design,

ought to have lasted for the required period despite its early collapse. The submission that

the works, as designed and built, were disconform to the 75 year requirement in the works

[258] Before the second element can be considered, it is necessary to decide precisely what

the contractor’s design was. The contract makes it clear that the design was something

which was to be “accepted” by the project manager (clause 21.2; Sch Pt 3, sec 1, Pt 6,

paras 6.2.8.1 and 6.2.8.4). What was accepted was the HRT RSM, (Drawing D152 etc) which

included Table 11, and Drawing D201, which detailed the rock support appropriate for

specified rock conditions. The contractor’s design was that contained in those documents,

each of which was accepted by Mr Sandilands. What occurred in the HRT as the TBM

progressed, with the preparation of the REC sheets and the mapping, was an exercise in the

implementation of that design. The REC sheets and mapping may have been “approved”

by Mr Sandilands, but they were not “accepted” as required by the contract. It is the

contract which must govern the relationship between the parties, irrespective of what those

in the tunnel thought they were doing in “approving” the defenders’ proposed tunnel

support. The later actings of the parties during the performance of the contract cannot assist

in its construction (Whitworth Street Estates v Miller [1970] AC 583, Lord Reid at 603). No

question of waiver or personal bar was raised. The contract makes it clear (cause 14.1), in

any event, that acceptance of a communication from the contractor or of his work does not

alter the contractor’s responsibility to provide the works or his liability for his design.

[259] In order to succeed in proving the existence of a defect on this second element, the

pursuers now rely solely on a contention that there was erodible rock that was not

shotcreted as it should have been in terms of Table 11; that being part of the design. The

commercial judge held (para [152]) that the collapse occurred because weak rock

deteriorated when submerged. Erodible rock was washed out and larger seams were

created. It may be assumed, from the judge’s finding, that some erodible rock had not been

shotcreted, although it does not follow that the erodibility of the rock was visible or

detectable by the geologists on site. The judge was careful to point out that the mechanism

of the collapse involved the deterioration of the rock, after it had been submerged, prior to it

becoming erodible. It should also not be assumed that whatever rock had become erodible

had not been shotcreted. The hazard of shotcrete cracking and being washed away was also

[260] The issue essentially boils down to a legal, and not a factual, question of whether

there was a mandatory requirement to shotcrete every area of erodible rock, however

unimportant the area might have seemed to an experienced geological engineer on site.

Although there may again have been a tendency on the part of the commercial judge to

interpret the contract terms by looking at how they were operated in practice, he rejected (at

para [167]) the contention that Table 11 had required the shotcreting of all erodible rock for

several reasons. One critical one was that, if it had done so, it ran counter to the other

contractual provisions which called for the exercise of engineering judgment on site. This is

a reference to: (a) clause 6.2.8.4 of Contractor’s Design (Sch Pt 3, sec 1), which stated that the

accepted design was to be developed through construction detailing and method

statements; (b) the provisions (ibid, sec 2 para 2.3.1) requiring the contractor to map the

tunnel after excavation and to agree the “rockmass parameters ... and the support class”

with the project manager; and (c) the Pöyry design statement (the Pöyry Guidelines) (Sch

Pt 7; Appdx 6, para 2.4) which required the contractor to determine the “actual support

requirements during the tunnel advance based on actual conditions encountered,

particularly with reference to the identified risk situations”. The latter was a reference to

one identified in the RSM, including the erodible rock hazard mentioned in Table 11.

[261] Neither Table 11, nor any other contractual document, stipulates that all erodible

rock required to be shotcreted. Table 11, as it was no doubt correctly described by

Prof Müller, was a “risk-handling matrix ... used ... during the tender negotiations”. It was

part of the RSM; being page 35 of some 38 pages. It comes at the end of a section of text

headed “6.5 Hazard control Tunnel/General Guidelines” which “suggests” methods of

dealing with hazards. It is a table which identifies a number of hazards, including the

detachment of wedges, cracking of shotcrete, scouring of the lining and “the erosion of

erodible rock during operation”. It specifies countermeasures; being shotcreting in the case

of identified erosion. However, it is tolerably clear, when the whole of the RSM is read, that

it is not prescriptive. It is, as it states, a guideline with suggested countermeasures to meet

predicted general hazards. It is not entirely clear that the specific hazard of erodible rock is

[262] The commercial judge’s approach to Table 11 is correct. It is simply not established,

from looking at the contractual documents, that all erodible rock required to be shotcreted.

In the circumstances, the case, based on an short passage in a table forming a relatively small

[263] It is only once it has been determined that there was a defect at take over (ie if the

above analysis is incorrect) that consideration requires to be given to the fourth question of

whether the limitation on the contractor’s liability for defects due to his design in: (i) the

description of the contractor’s main responsibilities; and (ii) the similar but more general

provisions of Option M, apply in a manner which removes any liability for defects if the

contractor proves that he used reasonable skill and care to ensure that his design complied

with the works information. Both of these provisions limit the liability of the contractor to

pay damages for loss which occurs as a result of a defect in the works due to the contractor’s

design. They affect liability, which is determined by section 8 of the contract (Risks and

Insurance). The effect of Option M is to qualify clause 80.1, by providing the contractor with

a defence of reasonable skill and care even when the loss or damage occurs because of a

defect which existed at take over. Clause 80.1 is the only provision in the contract which

apportions the risk of liability between the parties for a catastrophic failure of the type that

[264] The allocation of risk or liability for damages is separate from the question of who

should carry out work to correct defects and bear the initial cost of doing so. Section 4

(Testing and Defects) makes it clear that the contractor is obliged to correct a defect

(clause 40.4), irrespective of its cause during the defect notification period (clause 43.1). If

the contractor fails to do so, the employer can assess the cost of correction by another and

the contractor must pay this amount (clause 45.1). This provision allows an employer to

take payment for any remedial work, in lieu of the contractor carrying out the work, if the

[265] The obligation on the contractor to correct defects is qualified by the bespoke clauses,

including clause 46.1, which provides that he is not obliged to carry out repairs which are a

result of normal wear and tear, and clause 46.2, which provides for the defects date to be

delayed. In terms of clause 46.4, the parties agreed that the contractor would have an

obligation to correct defects “at his own cost and expense and with all possible speed” for a

period of 12 years after the issue of the defects certificate. If the contractor failed to perform

this obligation, the employer could recover the cost of employing others to carry out the

[266] Both the general and the bespoke provisions are silent on the question of who is

ultimately liable for the cost of the repairs. That is not surprising, given that section 4 deals

with “testing and defects”, and not the allocation of risk. Where an employer’s risk event

occurs, the contractor is obliged, under the terms of section 4, to return to site, to correct the

defect and to bear the cost of doing so. This is reinforced by the terms of clause 82.1, which

require the contractor to carry out repair works promptly, before the issue of the defects

certificate. If the contractor refuses to return, the employer may claim the cost of the repair

from the contractor. However, if the event is ultimately found to be an employer’s risk, the

contractor is entitled to be indemnified by the employer for the costs incurred as a

[267] The general description of the contractor’s main responsibilities provides that

liability for defects due to his design, which are not identified prior to the defects date, will

be limited so far as the contractor can prove that he used reasonable skill and care to ensure

that the design complied with the works information. This is consistent with Option M

which, when read with clause 80.1, provides that loss or damage to the works taken over by

the employer and occurring before the issue of the defects certificate, which is caused by a

defect which existed at take over, will be the contractor’s risk unless the contractor can

establish that he used reasonable skill and care. The defenders’ liability for identified and

unidentified defects, which give rise to loss or damage to the works, are both limited to the

[268] The structure of the contract is then that the contractor is obliged to correct all defects

notified prior to, or specified in, the defects certificate. This is regardless of whether the

defects were caused by fault and negligence in the design. This is clear from the provision

(clause 45.1) which permits the employer to recover the cost of having a defect corrected by

another from the contractor. Normally, the employer would only be at risk once the defects

certificate has been issued and then only for defects not specified in that certificate.

However, the bespoke provisions of this contract (clause 46.4) provided that the contractor

must make good at his own cost and expense any defect in the civil works appearing during

[269] Section 8 allocates liability to pay for loss or damage as a result of a defect in the

works following take over by the employer. That loss will be an employer’s risk, unless it

occurred before the issue of the defects certificate and was due to a defect which existed at

take over. In that event, the contractor will bear the liability for the loss and damage, unless

he can show that he exercised reasonable skill and care in ensuring that the design complied

with the works information. If the contractor is obliged to correct a defect which was an

employer’s risk event, he will be entitled to the benefit of the indemnity to recover his costs

[270] After the twelve year period, the defender is no longer obliged to correct defects.

During the period between take over and the issue of the defects certificate, any loss or

damage caused by the works is governed by clause 80.1. Following the issue of the defects

certificate, any loss or damage to the works is an employer’s risk event.

[271] On the hypothesis that the pursuers had demonstrated that the 75 year design

requirement had not been complied with, that would have to have been because of a defect

in the accepted design (presumably ultimately a failure to line the tunnel throughout the

CFZ). If they had proved that there had been a failure to shotcrete erodible rock, the failure

would have been one in implementing the accepted design (on the hypothesis that it

required this precaution). This would still fit into the description of a defect due to the

contractor’s “design”, which is unqualified by reference to the wording which defines a

defect. It is not suggested that those carrying out the support work in the HRT failed to

comply with Mr Taylor’s requirements. What he designated by way of HRT support is

properly classified as part of the contractor’s design when considering Option M.

Clause 11.2(15) refers to a defect occurring where a part of the works designed by the

contractor does not accord with the accepted design. The defenders thus escape liability for

a defect, even if it existed at take over, if they can prove (as in the event they did (infra)) that

[272] Although the point was not pressed, for completeness, the pre-watering up

inspections did not constitute an “event” within the meaning of clause 80.1. They were

[273] The incidence of the burden of proof is only significant if a judge at first instance is

unable to reach a concluded view on the facts, having heard the evidence (Thomas v Thomas

1947 SC (HL) 45, Lord Thankerton at 54). In the vast majority of cases, such a view can be

reached notwithstanding the absence of direct testimony on a particular matter. As has

often been said, onus seldom matters once the evidence is out (eg Salt International v Scottish

Ministers 2016 SLT 82, LJC (Carloway) at para [45]; Gibson v BICC & Co 1973 SC (HL) 15,

[274] One curious, albeit by no means unique, feature of this case is that what might

otherwise have been regarded as the crucial testimony on the issue of negligence, notably

but not exclusively that of Mr Taylor, was not adduced by the defenders, upon whom the

onus lay. There was no material produced to suggest that Mr Taylor could not be found and

duly cited to appear at court. It must be assumed that the decision of the defenders not to

call him was a deliberate one, even if the precise reason for that is not known to the court. If

the defenders had had any difficulty in tracing Mr Taylor, they could have sought the

court’s assistance, especially in circumstances in which the pursuers knew of his

[275] Where an incident occurs and the party upon whom the onus lies does not lead his

protagonist, there can be no doubt that the court can draw an inference adverse to that

party’s interests from that absence. The court would be entitled to find that the matter upon

which that person was peculiarly able to speak, in this case the exercise of due skill and care

in the HRT, had not been proved. The court could thereby hold that the other party’s case

was thus made out (see eg Binnie v Rederij Theodoro 1993 SC 71, LP (Hope) at 87). In this

context, there is considerable force in the pursuers’ submission that the absence of

Mr Taylor, and indeed the Jacobs and Pöyry geologists, ought to have been fatal to the

[276] However, as a generality, where there is some evidence pointing to a particular fact,

a judge will normally be expected to decide whether that fact has been proved in the

absence of an important, or even the central, witness or witnesses. Deciding a case on onus

is the exception rather than the rule. It ought to be regarded almost as a last resort when

some such evidence is present. A judge has a duty to decide the case on the evidence

adduced and to avoid doing so on the basis of onus if there is some evidence which does

support one or other or both of the parties’ contentions (see eg The Popi M [1985] 1 WLR 948,

Lord Brandon at 955). The issue in this case is whether, albeit that the commercial judge

would have been justified in deciding the case in favour of the pursuers on the basis of onus,

[277] The absence of Mr Taylor must certainly be regarded as significant. It could have

been regarded as determinative. However, there was evidence before the commercial judge

about what Mr Taylor had done, when assessing the geological conditions in the tunnel, and

about the care and attention which he had paid to his task. In particular, there were his REC

sheets and mappings, which were completed in some detail and spoken to by others. There

was, quantum valeat, the testimony from the TBM crew about the care which Mr Taylor took

and some evidence from several witnesses about his general level of skill and care. Most

significant, there was the fact that none of the other experienced engineering geologists, who

were in the HRT before its collapse, criticised Mr Taylor’s judgment on site. In addition,

there was testimony from several, albeit not all, of the experts to the effect that Mr Taylor’s

[278] The pursuers were able to point to questions arising from the form and content of the

documents produced which, they submitted, could only have been answered by Mr Taylor.

These will each be analysed on their individual merits (infra). Suffice it to say, this was not a

case in which the facts could not be determined in the absence of Mr Taylor. It was not a

case in which there was no evidence of what he had found and what he had decided. That

evidence came from the content of the documents produced and spoken to by witnesses.

There was testimony from which an inference could be drawn that reasonable skill and care

had been taken. Hills v Niksun [2016] IRLR 715 involved quite different circumstances in

which there was no evidence at all of the considerations upon which a decision had been

based. Here, there was evidence of what had, at least apparently, been seen and taken into

[279] The determination of whether the defenders exercised reasonable skill and care may

be classified as one of law. It is nevertheless heavily dependent upon findings of primary

fact in relation to: (i) the rock mass observed, or observable, in the HRT; (ii) the bases upon

which decisions regarding support in the HRT were to be made; and (iii) the decisions

[280] The court must bear in mind the limitation of its jurisdiction to reverse findings in

fact, having regard to the dicta in the trio of cases in the United Kingdom Supreme Court

(McGraddie v McGraddie 2014 SC (UKSC) 12; Henderson v Foxworth Investments 2014 SC

(UKSC) 203; and Royal Bank of Scotland v Carlyle 2015 SC (UKSC) 93), which heavily

influence how the traditional approach outlined in Thomas v Thomas 1947 SC (HL) 45 (Lord

[281] Findings of primary fact can only be interfered with by an appellate court where the

judge at first instance can be shown to have been “plainly wrong”. That much is clear. This

has been interpreted as meaning that the judge must have reached a decision which no

reasonable judge could have reached (Henderson v Foxworth Investments (supra) Lord Reed at

para 62) or that his decision cannot reasonably be explained or justified (ibid at para 67). As

“The court does not understand Lord Reed to be seeking to depart from the familiar

and long-settled approach of the Scottish courts hitherto in appeals on matters of

fact. Although some of the wording … might, if looked at in isolation, be taken to

suggest an approach redolent of the high test applicable in cases of judicial review, it

is clear from Lord Reed’s careful analysis of the dicta in Thomas v Thomas that what

he was doing was explaining in more modern language the meaning of ‘plainly

wrong’. This is consistent with his observation in McGraddie v McGraddie (para 5)

that: ‘While the law is not in doubt, its application has been inconsistent’”.

[282] Although the matter may have become one of some controversy in recent years at

least in the Inner House of the Court of Session, an appellate court can more easily reverse a

judge at first instance when what is under review is not a finding of primary fact but an

inference (or secondary fact) drawn from the primary fact or facts (cf Housen v Nikolaisen

[2002] 2 SCR 235; see also Royal Bank of Scotland v Carlyle 2014 SC 188, LJC (Carloway),

delivering the Opinion of the Court, at para [60]). It may do so with even greater confidence

if what is under consideration is the application of the law to the facts, whether primary or

inferential (see generally the analysis in “Appellate courts: Parts 1, 2 and 3” 2015 SLT (news)

[283] On the one hand, the advantages enjoyed by a judge at first instance in having seen

and heard the witnesses should not be underestimated. Furthermore, an appeal court must

have due regard to the limitations of the appeal process with its “[narrow focus] on

particular issues as opposed to viewing the case as a whole” (McGraddie v McGraddie (supra)

Lord Reed at para 33, citing Housen v Nikolaisen (supra) at para 14; see also Royal Bank of

Scotland v Carlyle (supra) Lord Hodge at para 22). On the other hand, the considerable

benefits of a three (or more) judge bench in terms of differences in view, varieties of legal

background and dialogue should not be overlooked (see Appellate courts: Parts 1, 2 and 3

(supra) at 127). The position of the Court of Session as a unitary court hearing reclaiming

motions from Outer House judges exercising delegated functions is also an important

consideration in assessing the scope of the peculiarly Scottish appellate jurisdiction (ibid).

The Inner House is not a separate appellate court as may be found in other jurisdictions.

[284] With these cautionary words borne firmly in mind, it is not possible to conclude that

the commercial judge has erred in finding that the defenders did exercise reasonable skill

[285] In the appellate process, the pursuers have founded upon very specific examples,

which they maintain demonstrate negligence. The pursuers’ submissions in relation to the

deficiencies, which were said to have existed in the manner in which the design was

implemented (or continued) in the HRT, have, once more, considerable force. It is incorrect

to state, as the commercial judge did, that none of those in the tunnel had recommended the

installation of a higher level of support at particular locations. At the time of the inspections

in January and February 2008, Pöyry had recommended additional support in certain areas.

Nevertheless, the judge was correct in his critical finding that no-one had identified a fault

that presented a threat to the HRT’s stability. The problems which had been identified by

Pöyry had been dealt with in the “closing off” of defects notice DN033, dated June 2008.

[286] Although Dr Palmström did agree with counsel that there had been a “missed

opportunity” in October 2008, when concerns had led to defects notice D041, this does not

advance matters, given that what was happening at that stage was not a study of the rock

mass by engineering geologists but an inspection to see that there was no dangerous debris

which might damage the turbine during watering up. Subject to what follows in relation to

the reliance on documents not spoken to by their authors, the judge was entitled to found

upon what Mr Fawcett had written in his report dated 16 April 2010 viz.: “There is no

recorded evidence of passing through any feature that would potentially cause the

catastrophic collapse that has occurred”. He was equally entitled to found upon Mr Speirs’

minute of 9 September 2007 that signs of the CFZ were imperceptible.

[287] The pursuers have undoubtedly pointed to features of the rock in the vicinity of the

collapse which, at least to a non-expert eye, may appear to provide strong clues about what

had occurred and why. The commercial judge’s assessment of the cause of the collapse is

not under significant challenge. Dr Wilhelm’s reconstruction in hindsight (supra) paints a

dramatic picture, which was broadly accepted by the judge, of submerged weak rock

deteriorating, erodible areas being washed out, opening up larger interconnected seams and

[288] The pursuers were able to point to what was recorded at chainages 2117-210110 about

the findings of kakirite and the references to particular angled shears in the blocked zone at

chainages 2121-2050. They founded upon to the ticks on the REC sheet for chainages 2084 to

208211. They were able to pose the question of why these areas were not provided with class

III or IV support given the terms of the RSM; notably Table 11. However potentially

objectionable the questions and inadequate the answers may have been, the defenders’

experts, notably Drs Büchi and Palmström, did provide a response to these questions. It

was, in essence, that, given Mr Taylor’s expertise, having specifically noted these features,

he could not have thought that they were of such significance as to require an increased

level of support. The judge was entitled to rely on this explanation and on the undisputed

fact that, at the time of excavation at least, none of the other engineering geologists from

Pöyry or Jacobs disagreed with Mr Taylor’s assessments of rock mass and support.

[289] The pursuers were able to question whether those inspecting the tunnel at the

excavation stage had considered Table 1112, given that it did not feature expressly on the

REC sheets13. They were able to suggest that the operation of the tunnel under hydraulic

pressure was not what Mr Taylor and his fellow geological engineers had in mind when

deciding upon the level of support. The evidence of Dr Broch at the Adjudication pointed to

this as a possibility, at least on a Norwegian model. Notwithstanding these points, the

commercial judge was entitled to hold that, the contention that operational circumstances

had not been considered during excavation, was a “startling proposition”. He was entitled

to rely on Prof Müller and Dr Wilhelm in this regard, since they had testified that those on

the ground had been thinking about the operation, and not just the excavation, of the HRT.

The terms of Mr Taylor’s email of August 2007 did not suggest otherwise. The judge was

entitled to reject the pursuers’ contention that what was being carried out was a two stage

process of, first, excavating a tunnel which was safe in plain air conditions and, only then,

considering separately operational concerns. No-one involved in the construction of the

tunnel spoke to such a plan and it was not reflected in the contract documents.

[290] The commercial judge did not require to make an express finding that Table 11 had

been considered, given his general findings about the approach of Mr Taylor and the other

engineering geologists on site. There was no reason to suppose that the geologists had been

unaware of the central documents in the works information and detailed design. Mr Taylor

must have been aware of them, given the form of the REC sheets, even if the wrong Table

was referred to in the template of the HRT sheet. The language used to identify the rock

class was consistent with awareness of the terms of the relevant documents. The term

“erodible rock” did not require a definition. It seems to have been generally understood.

The judge was correct to hold, as was indeed the case (supra) that Table 11 did not contain a

requirement that all erodible rock required to be shotcreted. The various experts were

generally, but not unanimously, agreed that the RSM was appropriate to determine rock

class. The judge was entitled to hold that what had been envisaged by the works

information and the accepted design was a method whereby an engineering judgment was

to be made at the face of the excavation in relation to support at particular locations.

[291] Given that this process was being overseen by both Pöyry and Jacobs, it is a

reasonable conclusion that what Mr Taylor was doing was making a decision which, at least

as the parties must have thought at the time, accorded with the contract. It did not involve

“painting by numbers”, or a “mechanistic” or “tick box” approach. Just because there was

some erodible rock or kakirite observed, or a shear of just under 500 recorded, did not mean

that 3600 shotcreting was automatic. The judge was entitled to pose the rhetorical question

of why the engineers should record rock conditions in such a detailed manner and then do

nothing about them; thus assuming the answer that they had taken all these matters into

account when making engineering judgments based on the totality of the rock mass

observations at the particular locations. The general absence of criticism of the engineering

judgments, once accepted, was all but a fatal blow to the case of negligence, and warranted

[292] The inspections in October and November 2008 had led to defects notices DN041 and

DN044. It is true that there were concerns about debris and voids in the invert, where rock

had been eroded, but the commercial judge did not consider that the problems identified

had involved major issues affecting stability, as distinct from matters which could be dealt

with at the stage of dewatering in due course. Again, given the absence of any engineering

geologist stating at the time that there were identifiable features which could lead to a

collapse, this was a finding which he was entitled to make, notwithstanding the power of

[293] The pursuers advanced an argument that it was possible that the defenders had not

appreciated exactly where the CFZ14 had been located on the HRT chainages. There may

have been a difference of opinion on this at one point, but Mr Taylor had made his own

views clear on where the CFZ would lie and what the related dangers for the TBM crew

were. His location was broadly accurate. The significant fact, and one which the

commercial judge was entitled to find on the evidence, was that there was no sign of the

CFZ, as the TBM progressed through its anticipated location, such that additional support

measures were required. The judge was entitled to find that the experienced TBM crew did

know what they were doing and could identify different rock classes, even if Dr Smith

thought that they might have been ignorant of the true conditions.

[294] The proposition that Pöyry had not examined the CFZ in January and February is

not supported by the evidence. Their inspections had led to the very substantial defects

notice D033 covering the relevant area. Even if they had failed to examine particular areas,

the evidence of Prof Broch made it clear that, on his inspection/walk-through of the HRT,

there was nothing significant which could have been found which would have predicted a

[295] In conclusion, it cannot be said that the commercial judge was “plainly wrong” in

concluding that the defenders had exercised reasonable care in relation to their design of the

HRT and in determining, at the face, how it was to be implemented. His decision can

reasonably be explained and justified on this central aspect of the proof.

[296] A judge requires to provide reasons for a judgment in a contested litigation in order

to demonstrate that justice has been done, and has been seen to have been done. The parties

are entitled to know how the judge has decided their case. This will normally involve

setting out the facts which have been admitted or proved and, where there has been a

dispute, the evidential basis for the facts found. The legal issues raised by the parties have

to be identified, usually in the form of a succinct summary of the submissions made. The

manner in which these issues have been resolved requires to be explained. This ought, if

possible, to be done in a concise manner which will be understandable, not just to the

lawyers involved in the litigation but also, at least, to the well informed reader of reasonable

intelligence, and the parties themselves. As was said in MacLeod’s Legal Representatives v

Highland Health Board 2016 SC 647 (Lord Brodie, delivering the Opinion of the Court, at

“the summarising and marshalling of points at an appropriate level of generality

with a view to focusing what is relevant to a resolution of the issues and not getting

[297] Regard must be had to the different potential readers; primarily the parties, but also

other interested persons, where the point is one of public and/or legal interest, and any

appellate court (ibid). A process of recording the evidence and submissions verbatim should

be avoided in favour of a considered selection of what is relevant, important, concise and

coherent. Although a judicial opinion will often require to be more structured and detailed

than those of administrative tribunals, the standard in the latter, which requires that the

reasons given must at least “leave the informed reader and the court in no real and

substantial doubt as to what the reasons for it were and what were the material

considerations which were taken into account in reaching it” (Wordie Property Co v Secretary

of State for Scotland 1984 SLT 345, LP (Emslie) at 348), is a valuable starting point for any

[298] The commercial judge adopted a style which ought to be regarded as a sound

example of how a written Opinion at first instance ought to be composed in a case of this

considerable magnitude. It must be borne firmly in mind that, although a judge’s reasoning

must always be adequate, the parties are entitled not only to that reasoning, but also to a

decision within a reasonable time. In this case, the judge made avizandum on 22 April 2016

and advised the case by the end of the year. Given that he would have had many other

cases to decide and duties to perform during that period, that was a remarkable

achievement where the parties had, as the judge observed, lodged 73,000 documents and

there were transcripts of testimony given over many months to be poured over. The judge

commented, not unreasonably, that it was unfortunate that the parties had been unable to

narrow the scope of the proof and at least to have lodged an agreed chronology and a list of

issues. The less parties assist the court in focusing the issues of fact and law, the more

difficult it is for the judge to reach a reasoned decision and to do so within a reasonable

[299] On the issues at proof, the commercial judge produced 72 pages of text, together

with an appendix containing drawings and photographs which illuminate his writing. He

set out, in limine, the structure of his Opinion in separate sections which make it clear to the

reader exactly what is being dealt with at the particular point. He identified the general

issues in an introduction. He provided a chronology of events, starting with the scheme and

going on to describe the contract, the construction, its collapse and the recovery project. He

had specific sections on why the tunnel collapsed, the witnesses and burden of proof,

whether (in relation to the works information or the design) there was a defect at takeover

and reasonable skill and care. He dealt separately with whether the defenders ought to have

returned to site, contributory negligence, causation, the costs of recovery and individual

heads of damages, before considering the counterclaim and ending with clear conclusions.

[300] In each of these sections the reasoning of the commercial judge is generally clear and

concise. It leaves the reader in no doubt about what his reasons for reaching a particular

view were. That is not to say that the reasoning is as clear as it might have been on all the

points raised, or that it might have been more expansive in certain areas, but it meets the

legal test. At no point does it take on the character of an oracular pronouncement (Dingley v

Chief Constable, Strathclyde Police 1998 SC 548, LP (Rodger) at 554 quoting from Davie v

Magistrates of Edinburgh 1953 SC 34, LP (Cooper) at 40). It should be borne in mind that the

proof was transcribed instantly by “Live-Note”. There was no need for a detailed rehearsal

of what all the witnesses had said, even if some detail has been necessary in the context of

the reclaiming motion, given the specific focused criticisms of the judge’s approach to

particular aspects of the evidence. The court has not found the absence of such a rehearsal a

(a) evidence shall not be excluded solely on the ground that it is hearsay;

(b) a statement made by a person otherwise than in the course of the proof shall

be admissible as evidence of any matter contained in the statement of which

proceedings, shall be entitled to find that fact proved by the evidence

“Statement” includes “any representation (however made or expressed) of fact or opinion”

[302] In terms of the joint minute, it was agreed that a number of specific witness

statements would be treated as “evidence in the cause”. It was then provided that:

“3. Save to the extent that evidence to the contrary has been heard or it is

(i) all productions and electronic documents are what they bear to be ...”.

[303] One of the purposes of the abolition of the prohibition against hearsay in civil

proceedings was to permit the use of documents as evidence in the absence of oral

testimony from their authors. It may be that, in certain circumstances, a judge will attach

little weight to a statement when the maker is available to testify, but that is a matter for

judgment. Hearsay, in the form of a document not spoken to by its author, is admissible as

proof of fact and it is for the judge at first instance to assess its weight. Where there is

evidence, oral or otherwise, to contradict it, the content may be rejected, but the opposite

may equally be the case. The hearsay may be accepted as more credible and reliable in the

[304] Where it is agreed that a document is what it bears to be, the content of the

document is evidence in causa. If the provenance of the document points to the accuracy of

the content, that content may be preferred to other evidence, including testimony. Thus,

where it is accepted that the REC sheets and the mapping were “what they bear to be” and

various witnesses spoke to the accuracy of Mr Taylor’s work, the content of these documents

can be accepted as true and accurate even in the absence of their author. Both parties appear

to have proceeded on that basis, at least for certain purposes. Similar considerations apply

to Mr Taylor’s emails and to Mr Speirs’ minute. The commercial judge could have decided

to discount the content of the minute, on the basis that Mr Speirs had not been called, but he

was not bound to do so and ought only to have done so if there was some reason to call its

[305] The report from Mr Fawcett falls into the same evidential category. It was open to

the commercial judge to take into account what an expert in tunnelling had reported, in the

context of the other evidence, and to use it to support the conclusions which he ultimately

reached. It was for the judge to determine whether the absence of its author, notably for the

purposes of cross-examination, was of significance in all the circumstances.

[306] In McEwan v Lothian Buses [2014] CSIH 12, there was a written statement from the

defenders’ bus driver which described the relevant road accident. It is not entirely clear

from the Opinion (paras [3]-[4]) whether this document was one of those which was agreed

by joint minute as being “what it bore to be”. The Opinion states (paras [4] and [18]) that

there was “no agreement ... about the evidential status of the document”, but if that were so,

the purpose of mentioning the terms of the joint minute would be obscure. If it had been so

agreed, it would have become a hearsay statement whose evidential value would have to

have been weighed in the balance and not automatically excluded. As it transpired

(para [18]), no one spoke to the provenance of the document and it is not surprising that it

was not regarded as material. That is not the position in this case.

[307] In any event, neither the content of Mr Speirs’ minute nor the conclusion of

Mr Fawcett’s report were contradicted by Dr Wilhelm’s ex post facto reconstruction of the

collapse. The absence of visible features indicative of a potential collapse is not inconsistent

with the occurrence of the collapse as a result of the unseen interlinking of shears above the

crown of the HRT. Equally, in relation to the evidence of Dr Broch and Prof Sloan and that

concerning the DAT, the weight to be given to the various adminicles, including the hearsay

provided by the experts, remained matters for the commercial judge to assess. Once again,

it cannot be affirmed that his assessment was “plainly wrong”.

[308] For reasons similar to those already given in relation to the commercial judge’s

findings on reasonable skill and care, he was entitled to hold that the use of such skill and

care would not have prevented the collapse. He found (para [260]) that only the installation

of class IV support would have done so. The pursuers had only said (paras [165] and [195])

that class III ought to have been provided in the exercise of such skill and care. As the

pursuers submitted, this may not matter in a situation in which the defenders were

ultimately liable to meet the costs of correcting any defect. However, for the reasons already

given, that is not what the contract said. The inescapable consequence is that, at least in so

far as the pursuers rely upon a breach of contract based on a failure to shotcrete erodible

rock, they have failed to prove that the collapse was caused by that breach and hence the

[309] Whether the secondaries were defects was primarily a matter of fact, albeit in a

contractual context. The commercial judge provided several reasons for concluding that the

areas identified by Prof Sloan were not defects. First, no-one had identified them as such in

2008. This was a point of some force. Any defects as were identified would have been

recorded in defects notice DN033, which was closed off, or DN041 or DN044 in November

and December 2008. Secondly, the judge accepted the evidence of Mr Becker that, when he

looked at the relevant areas some time after April 2010, the features could not be classified

as defects. Thirdly, he accepted the evidence of Prof Broch, who had looked at the areas in

January 2011, that the features were either rock falls, which could be left unsupported, or

expected minor rock falls. Fourthly, he accepted Dr Billig, who was more or less to the same

effect in relation to an absence of any threat to stability. The judge’s conclusion here can

[310] The pursuers maintain that they were entitled to adopt what might be described as a

precautionary approach to the features and argue that, if the areas required repair, they

ought to recover any cost of doing so. However, this claim is that the features constituted

defects requiring correction in terms of the contract. If the tunnel was stable, despite the

existence of these features, there was nothing which could constitute a defect in terms of the

[311] The ingenuity of relying on the defenders’ assessment of the cost of repair at

£30 million or thereby, and arguing that therefore the pursuers could claim the balance

between that and the sums paid to BAM as the measure of their loss by virtue of the

defenders’ failure to do the work in breach of clause 82.1, is immediately acknowledged, but

it must as quickly be dismissed. The evidence of Mr Hunter, which had vouched the

£30 million figure, was rejected by the commercial judge for reasons which were not

challenged. That left, as the only evidence of the costs of repair, that presented by the

pursuers. In the absence of any acceptable evidence that the costs would have been less

than those reasonably incurred by instructing BAM, the commercial judge was bound to

hold that such costs would have been the same, had the defenders themselves carried out

[312] Conversely, the commercial judge was entitled to the view that the pursuers had not

placed unlawful conditions on the defenders by insisting upon a concrete invert, the DAT

and a BPT. Each of these aspects of the recovery project was justified by the evidence. The

invert was a requirement of the Health and Safety Executive. The DAT was to yield benefits

in terms of ventilation, access and safety, even if it did not achieve the objective of driving

the BPT from both ends. The BPT had been agreed as the sensible solution, given the

obvious dangers involved in driving through the collapsed zone. The judge readily justified

[313] An award of expenses is a matter for discretionary judgment by the court of first

instance. Appeals, at least when dealing only with expenses, are “severely discouraged”

(Caldwell v Dykes (1906) 8F 839, LP (Dunedin) at 840) and entertained only in very

exceptional circumstances such as where, for example, some obvious miscarriage of justice

has occurred (Charles Rogers & Sons v G & H Mullen 1957 SLT 23, Lord Mackintosh at 25).

The court will only interfere where the court at first instances has taken into account

something which it ought not to have done, left a material matter out of account or reached

a decision which was “plainly wrong” (Ramm v Lothian and Borders Fire Board 1994 SC 226,

[314] One matter which does requires to be taken into account is the incidence of success,

given that the general rule is that expenses should follow the result. In a damages claim,

that is usually measured in a straightforward way by determining whether the pursuer has

succeeded in obtaining an award in his favour; a defender being in a position, if concerned

about expenses, to lodge an appropriate tender. Where, as here, the pursuers had secured a

substantial award of around £1.5 million, that ought normally to be used as a starting point

and the absence, as here, of any tender would be an important matter (ibid), albeit that an

award of expenses may be refused or modified if the successful party’s evidence or conduct

has been unsatisfactory (ibid at 228). The approach of the sheriff, rather than the sheriff

principal, in D Macdonald & Bros v Cosmos Decorators 1969 SLT (Sh Ct) 9 is the norm (see also

Howitt v Alexander & Sons 1948 SC 154 and William Nimmo & Co v Russell Construction & Co

(No. 2) 1997 SLT 122, Lord Morison, delivering the Opinion of the Second Division, at 122).

The court does not normally engage in an exercise of deducting portions of expenses in

accordance with its view of how much time was taken up in dealing with a point which

turned out to be of no value or effect (ibid). That approach may, however, not be followed

[315] Even when a party has been successful, in the manner described, expenses may be

refused or modified and a contra award may be appropriate in all the circumstances (Howitt

v Alexander & Sons (supra), LP (Cooper) at 158). In this case the commercial judge took the

view that almost the entire proof (and the case in general) had been taken up with matters

upon which the defenders were successful. It may be surprising that the judge made an

award almost entirely in the defenders’ favour, given that the pursuers had achieved some

success, not only in securing an award, but also in discreet areas; notably in resisting the

defenders’ own pleas of contributory negligence and in the counterclaim and quantification.

However, the judge was undoubtedly in the best position to assess those matters (Ramm v

Lothian and Borders Fire Board (supra) at 229). Even if it may be said that the judge could have

expressed it in more expansive terms, the reasoning which he gives, and which is broadly

consistent with the nature of the pleadings and the extent of the proof, is readily

understandable and justifiable. The deduction of 5% may seem very small in relative terms

(representing only about 2½% of the total expenses of both parties) but, as was submitted, it

[316] Having regard to all the circumstances, there is no material error in the commercial

[317] The obligation was for the defenders to provide insurance in the joint names of the

parties in respect of (amongst other things) loss of, or damage to, the works caused by

events which were at their risk prior to the issue of the defects certificate (clause 84). All

risks, which were not those of the pursuers (typically claims due to the pursuers’ fault or

negligence (clause 80.1)), were carried by the defenders (clause 81). The insurance would

encompass damage to the tunnel, whether caused by the contractors, third parties or other

external factors. In the same part of the contract as the provision regarding insurance was

that relating to indemnity. Each party required to indemnify the other against claims due to

an event which was at that party’s risk (clause 83.1), with the liability reduced if the events

[318] If the tunnel collapse had been due to a breach of the contract on the part of the

defenders, the pursuers’ argument was that they could sue the defenders for the resultant

loss, even although the defenders may have insured themselves and the pursuers in respect

of that loss. The practical consequence of this could be that the pursuers would recover a

sum equivalent to that which their insurers had already paid out (or were liable to pay) to

them and which they would presumably be obliged to return. The circularity of this does

not pose a problem in practical terms if the basic principle that insurance is res inter alios acta

were to be applied. Any pointless litigation could be marked by an appropriate award of

expenses. The insurmountable difficulty with this straightforward approach, however, is

the highly persuasive authorities which dictate otherwise where the insurance is joint.

[319] The UK Supreme Court decision in the “Ocean Victory” [2017] 1 WLR 1793 was

decided after the commercial judge’s Opinion. It refused the appeal from the judgment of

the Court of Appeal, which the judge had considered. Although the judgments of the UK

Supreme Court are obiter on this point, it was not contended otherwise than that a joint

insurance clause precluded an action for damages covered by that insurance. Both the

majority and the minority considered that this was because, as a matter of construction,

there was a resultant implied term in the contract to that effect (Lord Toulson at paras 139-

[320] The commercial judge’s view (para [81]), that there was no reason in this case to give

primacy to the insurance clause (clause 84) over the indemnity provision (clause 83), is

correct. Taking out a joint insurance policy had utility and was prudent, but there was no

reason to suppose that it would supplant liability. The presumption in CRS (supra, Lord

Hope at para 65), that there is an implied term in a contract preventing litigation between

the parties, where there is joint insurance, can be rebutted having regard to the terms of the

contract. That is the situation with this contract. There was certainly no necessity to imply

such a term for the purposes of business efficacy. CRS (supra) turned, at least in part, on the

use of the term “liable” in section 1 of the Civil Liability (Contribution) Act 1978 (not

applicable to Scotland) and on the fact that the liability on the contractors had been

expressly excluded. The situation is different when the contract envisages that one co-

insured may be liable to the other in damages within the cover provided by the policy (Tyco

Fire v Rolls-Royce Motor Cars [2008] Lloyd’s LR 617, Rix LJ at para 77).

[321] Clause 82.1 required the defenders to replace the loss of, and to repair damage to, the

works. It did not restrict itself to loss or damage due to a defect as defined. It is an

obligation to replace and repair, whatever the reason for the loss or damage. There can be

little doubt that the collapse of the tunnel resulted in loss or damage to the works. The

defenders were bound to replace or repair the damage. They did not do so; indeed they

refused to do so and were thereby in breach of this important contractual provision. That

breach occurred irrespective of where the risk fell, in terms of clause 80.1, so far as the

provisions for indemnity (clause 83.1), and hence payment, were concerned.

[322] The defenders’ counterclaim is predicated on the proposition that they were entitled

to carry out the works and that the pursuers had prevented them from doing so. However,

from the commercial judge’s findings in fact, it cannot be concluded that this was so. The

pursuers had instructed the defenders to carry out remedial works by letter dated 31 August

2009 under reference to defects notice DN053. The response from the defenders, in their

letter of 25 September 2009, had been that they were “ready” to repair the collapse, as they

had recognised that they were required to do so in terms of clause 82.1. However, it was the

defenders who had imposed a condition, which was not stipulated in the contract, that the

pursuers should agree to pay for the remedial works in advance; ie before the issue of risk

could have been determined. Furthermore, they had insisted in their letter of 12 November

2009 that they should be paid on a cost reimbursable basis. They refused to proceed on an

[323] In these circumstances, the commercial judge correctly determined that the

defenders could not found on clause 82.1 to claim loss of profits in circumstances in which

they had refused to perform their obligation under that same clause.

[324] The claim for loss of profit fails for the reasons given in relation to mutuality of

obligations under clause 82.1. The defenders had refused to carry out the repair work and

[325] The indemnity provision of clause 83.1 relates inter alia to costs due to an event

which is at the party’s risk. In this case the risk event was loss of, or damage to, the works

under clause 80.1 (not falling within the exception). The question becomes one of whether

the costs claimed are due to the tunnel’s collapse. It is not disputed that the collapse has to

be the “effective”, as distinct from a “but for”, cause (ENE Kos I v Petroleo Brasileiro (No. 2)

[326] The defenders’ monitoring of the BAM works was not, as the commercial judge

correctly held, effectively caused by the tunnel’s collapse. The defenders did not require to

monitor the works because of the collapse. They elected to do this for their own purposes;

no doubt with an eye to potential litigation. Similarly, the defenders’ decisions to refer some

of the issues to two adjudications were not caused by the collapse. They were voluntary acts

made in order to secure payment of sums, of the type now claimed, and to obtain a

provisional view on liability. The defenders’ expenses in this litigation are not effectively

caused by the collapse but by the pursuers’ act in deciding to sue them.

[327] For the reasons essentially given, albeit in short form, by the commercial judge, the

[328] If an adjudicator has made an award, which is later determined by the court to have

been wrong, it follows that the sum must be repaid (Aspect Contracts v Higgins Construction

[2015] 1 WLR 2961, Lord Mance at paras 23-24). That is the position here where the court

has determined that the sums awarded under clause 83 ought not to have been awarded

[329] Clause 48 provided that, if the scheme was not operating after completion, the

contractor was required to pay certain sums to compensate for “low availability” up to a

ceiling of £1 million. As the commercial judge determined, this sum was due under this

clause. However, if it was caused by the tunnel collapse, the pursuers would be liable to

compensate the defenders for this under clause 83.1. The delay, or at least a significant part

of it, which produced these damages, was not caused by the collapse but by the defenders’

failure, in breach of their obligation under clause 82.1, to proceed “promptly” with the

recovery project. As was submitted, the investigations ought to have been completed by

mid-December 2009 but they had not even commenced until then. There may not have been

a precise calculation of the damage sustained as a result of the delay but, as the judge held, it

[330] The commercial judge explained that, in his assessment of damages, he had adopted

the approach taken in Banco de Portugal v Waterlow & Sons 1932 AC 452 (Lord Macmillan at

“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

It does not now appear to be disputed that the pursuers did pay the sums claimed to BAM

as part of the recovery project. There is no reason to suppose that the pursuers were other

than business-like in their approach to the contract with BAM, albeit against a background

of wishing to ensure that the tunnel did not collapse again. The judge held, as a matter of

fact, that the pursuers had acted on the basis of professional advice and had instituted an

exacting costs monitoring regime. The BAM contract, which was under Option E, was the

only basis upon which the defenders themselves would have contracted. In these

circumstances, it is not possible to fault the judge’s conclusions on reasonableness.

[331] The pursuers required to establish, not “causation”, but that what they had done had

been reasonably required to put them back in the same situation as if the collapse had not

occurred. That was established in terms of the commercial judge’s general findings on how

the pursuers had approached the recovery works. It was sufficient to prove the actual costs

of these works, especially given that the judge had also found that BAM’s profit margin had

[332] The decision to complete the works with dispatch was one aspect of the pursuers’

decision making which the commercial judge held was reasonable in the circumstances.

Whether or not loss of profit was allowable in terms of the contract, the pursuers were

entitled to proceed in a manner which kept their on-going losses to a reasonable minimum.

It may be that the defenders, and their witnesses, were of the view that the DAT was

unnecessary, but the judge was entitled to the view that it formed part of the necessary

remedial works and brought benefits in terms of ventilation, access and safety. Its

construction did not fall under the £1 million cap in clause 48.1.

[333] The absence of operational notes dealing with the needle valve opening parameters

did not amount to a fault (or negligence) on the part of the pursuers. The obligation to

explain how to operate the scheme rested on the defenders but, apparently, none of the

52 files provided by the defenders referred to needle valve openings.

[334] The failure to act upon the needle opening readings cannot be classified as a fault

either. The experts on site, namely Andritz, had not reported any problem requiring urgent

action. The commercial judge, having heard the evidence of this matter from experts on

both sides, had preferred the testimony of the pursuers’ experts that it was understandable

that no action would be taken on the unusual readings given that testing was ongoing and

the scheme returned to full output after the unusual readings occurred. It cannot be said

[335] In the result, the court ought to refuse the reclaiming motion and the cross appeal

and adhere to the commercial judge’s interlocutor of 2 February 2017.

Pursuers and Reclaimers: Moynihan QC, Barne QC; CMS Cameron McKenna Nabarro

Defenders and Respondents: McBrearty QC, Richardson QC; Clyde & Co.

[336] I have had the considerable advantage of having read in draft the opinions of both

your Lordship in the chair and Lord Glennie. This should enable me to address the difficult

issues in this case with much greater brevity than would otherwise have been the case. I am

particularly grateful to your Lordship in the chair for your clear and accurate summary of

the contractual terms, the factual background, the evidence before the commercial judge and

the submissions before this court, which I adopt in its entirety.

[337] The power station and associated tunnels, and the headrace (HRT) and tailrace (TRT)

tunnels, were taken over by the employer on 18 December 2008. The defects date was stated

to be 104 weeks after completion date. By 12 April 2009 a catastrophic collapse had occurred

in a substantial section of the HRT, which resulted in the closure of the power station for a

protracted period, and significant and costly remedial works. The power station did not

[338] The collapse therefore occurred about four months after take over, and well before

the expiry of the defects period. In order to answer the question “Where does the risk for

the collapse lie?”, it is necessary to look in the first place to the terms of clause 80.1, which

specifies the Employer’s risks. The fourth main bullet point of this clause includes among

“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

It follows that the collapse of a part of the HRT after take over by the Employer would be an

Employer’s risk unless it was due to a Defect which existed at take over.

 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

 the Contractor’s design which has been accepted by the Project Manager.”

The pursuers argue that the collapse of part of the HRT was due to a Defect which existed at

take over, under two limbs – (1) that this was a part of the works which was not in

accordance with the Works Information, and (2) that it was a part of the works designed by

the Contractor which was not in accordance with the Contractor’s design which had been

[340] The HRT was a principal component of the civil works. The Works Information

provided (at para 6.3.2), under the heading “Design Life”, that it should provide reliable

service without requirement for major refurbishment or significant capital expenditure for

75 years. It collapsed after four months. The commercial judge found (at para [147]) that the

short answer to the question “Why did the tunnel collapse?” was that there was not enough

support: poor rock conditions coincided with insufficient shotcrete and rock bolts. The

commercial judge gave further detail at para [152]. The simple argument for the pursuers

was that this amounted to a Defect because it was a part of the works which was not in

accordance with the Works Information. There was no suggestion, either in evidence before

the commercial judge or in submissions to this court, that there was any intervening event

after take over which might have caused or contributed to the collapse. The Defect, namely

the non-conformity with the Works Information, must have existed at take over. The

collapse in the HRT is therefore excluded from the Employer’s risks in terms of clause 80.1,

[341] I am persuaded that the pursuers’ analysis in this respect is correct, and that the

collapse of part of the HRT is indeed a Contractor’s risk. I have been hesitant to reach this

conclusion, partly because of the time spent in evidence and submissions before the

commercial judge, and in arguments before this court, on other issues and alternative

arguments. Can the point be as simple as this? I am mindful of Lord Glennie’s observation

that the question has to be asked, and must be answered, not in the real world but in the

world of contractual interpretation; and I am equally mindful of all the recent guidance as to

the correct approach to the interpretation of commercial contracts – eg Arnold v Britton

[2015] AC 1619, @SIPP Pension Trustees v Insight Travel Services 2016 SC 243, Hoe International

v Andersen 2017 SC 313, Rainy Sky v Kookmin Bank [2011] 1 WLR 2900, and Wood v Capita

Insurance Services [2017] AC 1173, to name a few. However, I am not persuaded that the real

world and the world of contractual interpretation are always necessarily to be treated as

parallel universes. Where a contract provides that the principal components of a scheme

will provide reliable service without requirement for major refurbishment or significant

capital expenditure for 75 years, and one of those components suffers a catastrophic collapse

within about four months of take over requiring significant capital expenditure, the

application of the principles of contractual interpretation may result in the same answer as

[342] I do not suggest that para 6.3.2 provided a warranty that the HRT would in fact last

for 75 years without major refurbishment or significant capital expenditure. The Employer

had the benefit of the 104 week defects period from take over to determine whether the HRT

did have a design life of 75 years. That, it seems to me, is consistent with the approach taken

by the UK Supreme Court in MT Højgaard v E.ON Climate & Renewables UK Robin Rigg East

[2017] UKSC 59. The assessment of whether the component had the requisite design life fell

to be made at the defects date, and did not require to be made before then. By that date, it

was abundantly clear that the HRT did not have a design life of 75 years, as it had already

[343] It is, I think, important that the assessment falls to be made at the defects date, and

not before. Your Lordship in the chair appears to rely (at paras [255] to [257]) on the general

consensus of the experts that, at the stage of design (or perhaps on the basis of what they

saw before take over, or on the basis of the ground conditions understood to exist at the time

of the design) that the design ought to have lasted for 75 years. That involves the exercise of

foresight and judgment in assessing what may occur in the future. In the present case, no

foresight is required; at the defects date one has the benefit of hindsight. The HRT had

[344] Lord Glennie has given a fuller analysis of this limb of the pursuers’ argument; I find

myself in complete agreement with his analysis and conclusion on it. I agree that the loss or

damage to the works by virtue of the collapse of a portion of the HRT was at the

Limb two – not in accordance with the Contractor’s design which has been accepted by

[345] It must be kept firmly in mind that in order to constitute a Defect under this limb

there must be a disconformity between the part of the works being considered and the

Contractor’s design which has been accepted by the Project Manager (emphasis added). I

am in complete agreement with your Lordship in the chair (at para [258] above) that the

Contractor’s design which was accepted by the Project Manager was the HRT RSM, which

included Table 11, and that what occurred in the HRT as the TBM progressed, with the

preparation of the REC sheets and mapping, was an exercise in the implementation of that

design. The distinction between “approval” by Mr Sandilands and “acceptance” for the

[346] One of the hazards identified in Table 11 for the HRT was “erosion of erodible rock

during operation”. The Table assessed the level of risk before and after mitigation

measures. This hazard was assessed as having a risk rating of 9 and to be a high risk before

mitigation measures. The countermeasures specified to deal with the hazard were

“application of shotcrete if not already covered/’protected’ by steel rib support”. The risk

after the specified countermeasures had been implemented was reduced to 3, and was

assessed as being low. This was the relevant part of the Contractor’s design which was

approved by the Project Manager. It obliged the Contractor to apply shotcrete to erodible

rock if this was not already covered or protected by steel rib support. The Contractor did

not fulfil this obligation. The commercial judge found that the HRT collapsed because there

was not enough support: poor rock conditions coincided with insufficient shotcrete and rock

bolts (paras [147] and [152]). As your Lordship in the chair observes (at para [259]) it may be

assumed from the commercial judge’s findings that some erodible rock had not been

[347] The question of whether the presence of the erodible rock ought to have been

detected by those operating the TBM, or by the geologists on site at the time of excavation, is

not in my view relevant at this stage. It may be relevant to the issue of Option M and

whether this affects the Contractor’s liability (on which see below), but it is not relevant to

the issue of whether a “limb two defect” existed. The Contractor’s design which had been

accepted by the Project Manager obliged the Contractor to apply shotcrete to erodible rock if

this was not already covered or protected by steel rib support, and it is clear from the risk

ratings in Table 11 that this was an important element of that design. It was not fulfilled.

The HRT as implemented was therefore a part of the works designed by the Contractor

which was not in accordance with the Contractor’s design which had been accepted by the

Project Manager. I agree with Lord Glennie that the pursuers have made out their limb two

argument, and that on this basis too the loss or damage to the works by virtue of the

collapse of a portion of the HRT was at the Contractor’s risk. I adopt Lord Glennie’s

reasoning in relation to limb two, and in particular his treatment of points (a) – (e) in

paragraph [166] of the commercial judge’s Opinion, set out at paragraphs [382] to [388] of

[348] If Option M applies, it provides a limitation on the Contractor’s liability for Defects

in the works due to his design, so far as he proves that he used reasonable skill and care to

[349] As discussed above, the cause of the collapse was found by the commercial judge to

have been insufficient shotcrete and rock bolts being provided where there were poor rock

conditions. This was part of the implementation of the design, not part of the design itself.

To this extent, I agree with your Lordship in the chair at paragraph [258]. However, your

Lordship goes on to express the view (at para [271]) that what Mr Taylor designated by way

of HRT support is properly classified as part of the Contractor’s design when considering

Option M. I do not agree with this. Option M is concerned with Defects – that is, Defects as

defined in clause 11.2(15). As discussed with regard to limb two above, it is important to

keep in mind that in order to constitute a Defect there must be a disconformity between the

part of the works being considered and the Contractor’s design which has been accepted by

the Project Manager. What Mr Taylor designated by way of HRT support was never

accepted by the Project Manager for the purpose of clause 21.2. Although Mr Sandilands

may have indicated his informal approval, this could not amount to formal acceptance for

this purpose. Anything arising from Mr Taylor’s designation by way of HRT support was

part of the implementation of the design, but it was not part of the design accepted by the

Project Manager. It was therefore not a Defect in the works due to the Contractor’s design

covered by Option M, and the limitation of liability provided by Option M does not apply to

[350] The finding by the commercial judge that there was insufficient shotcrete and rock

bolts provided where there were poor rock conditions places the cause of the collapse as a

failure of implementation. Whilst this might be categorised as a “defect” in ordinary

parlance, it is not a Defect in the works due to the Contractor’s design, so Option M does not

[351] I agree with everything that your Lordship in the chair says at paragraphs [273] to

[278] above. I have nothing to add, except to observe for completeness that the decision of

the Second Division in HBJ Claimants v Glasgow City Council 2017 SLT 1135, in so far as

relating to the burden of proof, depended on the particular circumstances of that case, in

which no expert evidence was led by the employers and the tribunal reached a conclusion

having complained that it did not have sufficient evidence to enable it to do so. In an

unusual case such as that, the burden of proof may remain of importance, but it is not of

significance in the present case. The fact that Mr Taylor did not give evidence was perhaps

surprising, and was significant, but there was sufficient evidence from other sources to

enable the facts to be determined in the absence of Mr Taylor’s evidence.

[352] On the view I have taken, it is not necessary to address the issue of reasonable skill

and care. However, I find myself in complete agreement with the views expressed by your

Lordship in the chair at paragraphs [279] to [295]. I agree that it is not possible to conclude

that the commercial judge erred in finding that the defenders did exercise reasonable skill

and care. However, for the reasons given above, I do not consider that this has any

[353] I am in complete agreement with your Lordship in the chair in respect of each of

[354] I regret that I disagree with the views expressed by your Lordship in the chair

regarding causation at paragraph [308]. The commercial judge dealt with causation shortly,

at paragraph [195] of his Opinion. His reasoning has some relevance to a claim based on a

failure to exercise reasonable skill and care, but I do not consider that it is applicable to a

situation where the collapse was due to a Defect which existed at take over, either under

limb one or limb two. I agree with Lord Glennie (at para [376]) that no causation point

arises in respect of a limb one defect, or, if it does, it is readily answered in favour of the

pursuers. With regard to limb two, I agree with Lord Glennie’s reasoning at

paragraph [390]. The cause of the collapse was the coincidence of poor rock conditions with

insufficient shotcrete and rock bolts. I consider that the pursuers’ pleadings were sufficient

to cover this, and I am not persuaded that their claim founders on the rock of causation.

[355] I am in complete agreement with everything that your Lordship in the chair says at

paragraphs [309] to [316], and there is nothing useful that I can add in relation to these

[356] Both your Lordship in the chair and Lord Glennie reject the defenders’ cross appeal

on this point. I agree. I am persuaded by Lord Glennie’s careful reasoning on this point,

and for the reasons which he gives I consider that the joint insurance point does not assist

[357] For the reasons which your Lordship in the Chair gives at paragraphs [321] to [323] I

agree that there is no merit in the defenders’ arguments in this regard.

[358] I am in complete agreement with the views expressed by your Lordship in the chair

in relation to the remaining issues, namely the counterclaim (paras [324] – [327] ), repayment

of the second adjudicator’s award (para [328] ), low availability damages (para [329]),

quantum (and DAT) (at paras [330] – [332] ), and the operation of the scheme (paras [333/4]).

There is nothing that I wish to add in respect of any of these matters.

[359] For these reasons I consider that the reclaiming motion should be allowed and the

cross appeal refused. I see no reason to depart from the approach to quantification taken by

the commercial judge in his Note dated 31 January 2017. I would consider that an award

(1) £107,617,830.94 in respect of costs of the recovery project, as set out by the

(3) £32,357.98 in respect of part of the adjudicator’s award; and

(4) €388,720.27, also in respect of part of the adjudicator’s award.

For the reasons explained by the commercial judge at para 10 of his Note, interest should be

awarded at the rate of 4% per annum from the date of citation until the date of decree

pronounced by this court in respect of head 1, and thereafter at the rate of 8% until payment;

and at the rate of 4% from the date of citation until 2 February 2017 in respect of heads 2, 3

and 4, and thereafter at the rate of 8% until payment. I would reserve all questions of

expenses, both before the commercial judge and before this court.

Pursuers and Reclaimers: Moynihan QC, Barne QC; CMS Cameron McKenna Nabarro

Defenders and Respondents: McBrearty QC, Richardson QC; Clyde & Co.

[360] I am grateful to your Lordship in the chair for having set out so comprehensively the

facts, the evidence and the arguments pertaining to this Reclaiming Motion, as well as the

relevant contract terms. I readily adopt those parts of your Lordship’s Opinion.

[361] I find myself in agreement with your Lordship on most of the issues arising for

decision by this court. In particular I agree with your Lordship’s resolution of the issue

which was at the centre of the dispute between the parties, namely the impact of the

decision to include Option M. In the course of the hearing before this court both parties

presented their arguments on this issue on the premise that there was a potential conflict

between, on the one hand, a strict liability regime contained within clauses 43 and 46.4

(absolute obligation to correct defects notified or discovered by certain dates) and clause

82.1 (absolute obligation to replace or repair loss or damage to the works occurring before

the defects date); and, on the other, the proviso in Option M to the effect that the contractor

was not liable for Defects in the works due to his design if he could prove that he used

“reasonable skill and care” to ensure that his design complied with the Works Information.

As your Lordship demonstrates, there is no such conflict. The obligation to correct defects

(clauses 43 and 46.4) and the obligation to replace or repair loss or damage to the works

(clause 82.1), and in each case the obligation, in default of so doing, to pay the employer the

cost of him having done the work himself, is essentially a practical means of ensuring that

when defects are identified or loss or damage occurs they are immediately rectified by the

contractor and/or at his expense. The ultimate liability for such defects, loss and/or damage,

however, is determined in accordance with the allocation of risk under clauses 80.1 and 81.1;

and the effect of incorporating Option M is to convert the contractor’s liability for the

occurrences which are at his risk from one of strict liability to one which is qualified by

[362] However, I differ from your Lordship on the question whether the collapse of the

tunnel was due to a Defect which existed at takeover. I consider that the collapse of the

tunnel was indeed due to a Defect existing at takeover. Further, I consider that that Defect

was not due to the contractor’s design of the works but rather to the implementation of that

design. In those circumstances Option M is not engaged, and the defence of having used

reasonable skill and care to ensure that the design complied with the Works Information is

not available to the contractor. It follows from this, in my opinion, that the collapse of the

tunnel was a contractor’s risk in terms of section 81.1 of the contract; and the defenders are

liable to the pursuers for the costs of repairing the tunnel. On that basis I would allow the

[363] In what follows I shall explain why I have come to that view. In addition I propose

to say a few words about the joint insurance point on which, though I agree on the outcome,

[364] I start by putting the issue in some sort of context. The question whether the collapse

of the tunnel was due to a Defect arises because of the terms of section 8 of the contract,

“Risks and insurance”. Clause 80.1 defines “Employer’s risks” under a number of bullet

points and sub-bullet points. The reference to a Defect in this connection is in the fourth

main bullet point. Read short, with a focus on this particular part of the clause, clause 80.1

 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

Clause 81.1 deals with contractor’s risks and provides that from the start of the works until

the Defects Certificate has been issued, risks which are not carried by the employer are

[365] The works, including the tunnel, were taken over by the pursuers in December 2008.

The take over certificate was issued on 18 December. The tunnel began to collapse shortly

thereafter, and finally collapsed within about six months of that date, that is to say some six

months into the period of about two years between take over and the issue of the Defects

Certificate. The collapse happened, and the loss or damage occurred, therefore, within the

period covered by that provision. And since no work had been done on the tunnel since it

was taken over by the pursuers, it must follow that whatever the problem was which caused

the tunnel to collapse, that problem already existed at take over, even though it took a few

[366] But the question which has to be asked for the purposes of the fourth main bullet

point under clause 80.1 is this: was the collapse due to a “Defect”? If not, that part of clause

80.1, which acts by way of an exception to employer’s risk, is not engaged; and there would

be no doubt that the collapse was an employer’s risk event. At first blush, and in the real

world, to ask whether the collapse was due to a defect might sound like a silly question. If a

tunnel costing well over £100m and designed to last for 75 years or thereabouts collapses

within the first six months after hand over, there was obviously something wrong with it; it

was defective; there was a defect. How could there be any doubt about it? But the question

has to be asked, and must be answered, not in the real world, but in the world of contractual

interpretation; and, specifically, in the context of this contract, in which the term “Defect” is

[367] The definition of Defect is to be found in Section 1, “General”. Clause 11 sets out a

list of “Identified and defined terms” and gives their meanings. Defined terms are given

capital initials. “A Defect” is a defined term and its meaning is given in clause 11.2(15):

 a part of the works which is not in accordance with the Works

 a part of the works designed by the Contractor which is not in accordance

 the Contractor’s design which has been accepted by the Project

The first bullet point under this definition (“a part of the works which is not in accordance

with the Works Information”) was, for convenience, referred to in argument as a limb one

defect (or a first limb defect), while the second was referred to as a limb two (or second limb)

[368] The contract comprises a number of documents listed in paragraph (1) of the Form of

Contract. In addition to the Form of Contract itself, and Schedule Part 1 (the “Contract

Data”, containing both the bespoke and incorporated numbered contract terms), those

documents include Schedule Part 3 (Works Information) and Schedule Part 7 (Works

Information for the Contractor’s design). Schedule Part 3 (Works Information) is the “Works

Information” referred to in the first main bullet point under clause 11.2(15). It may be that

Schedule Part 7 is also part of that “Works Information”, but nothing turns on this since I

agree with your Lordship (in para [253]) that nothing in Schedule Part 7 qualifies the

relevant part of the Works Information in Schedule Part 3. Schedule Part 3 (Works

Information) is divided into a number of sections. Within Section 1 (“General”), part 2

contains a general description of the works. Para 2.1 includes, among the employer’s

objectives for the development of the scheme, “reliable service without the requirement for

major refurbishment or significant capital expenditure within the design life of the scheme”.

Part 6 deals with “Contractor’s Design” and covers matters such as design responsibility,

design procedures and design development. Within para 6.3 (headed “Design Brief”), para

6.3.1 requires the contractor to design a hydro scheme with a minimum guaranteed

performance according to certain Performance Criteria detailed within the Contract Data,

while para 6.3.3 sets out the specifications to be followed in respect of plant, materials,

equipment and practices. Of particular relevance here is para 6.3.2, headed “Design Life”,

The design life of the principal components associated with the scheme are detailed

below. The components provide reliable service without requirement for major

refurbishment or significant capital expenditure for a period of time as follows:

It was not in dispute between the parties that the Head Race Tunnel (“HRT”) was one of the

[369] It is convenient at this point to consider the meaning and effect of such a provision.

One question is whether it is to be read as a 75 year warranty for the Civil Works. That

question was discussed in MT Højgaard v E. ON Climate & Renewables UK Robin Rigg East

[2017] UKSC 59 which concerned a contractual requirement that wind turbines had a design

life of 20 years. As in the present case, there was a two-year period after hand over within

which the turbines could be inspected and defects detected. Lord Neuberger, with whom

the other Justices agreed, considered (at para 32) that it was unnecessary to decide whether

that requirement amounted to a warranty that the foundations would have a lifetime of 20

years. His preferred interpretation was that the 20 year design life requirement “did not

guarantee that the foundations would last 20 years without replacement, but that they had

been designed to last for 20 years without replacement” (see para 30). Such an

interpretation meant that the contractor complied with his obligations if he handed over the

works in such a condition; and the employer had the whole of the two year defects period

within which to determine whether the Works did in fact have a 20 year design life. Mr

Moynihan QC, for the pursuers, adopted that approach. He did not argue that para 6.3.2

imported a warranty or guarantee that the HRT would last 75 years without major

refurbishment or significant capital expenditure. But he did argue, consistently with the

approach in MT Højgaard, that para 6.3.2 was an assurance that the tunnel would be handed

over with a design life of 75 years. That assurance was self-evidently not satisfied, since

within the two year period after hand over the tunnel had completely collapsed.

[370] There is one further point to make in respect of para 6.3.2 of the Works Information

in Schedule Part 3. There appears to be no universally recognised definition of “design life”,

though a number of law and business dictionaries suggest that it refers to the period of time

during which the relevant part of the works (in this case the tunnel) can be expected to

function properly without major repairs. In the present case it is not necessary to search in

the abstract for a definition of design life, since para 6.3.2 provides its own definition:

(reading from the second sentence) the component (ie the tunnel) is to provide “reliable

service without requirement for major refurbishment or significant capital expenditure for a

period of time as [specified]”, ie for 75 years. This repeats the wording in para 2.1

(Employer’s Objective) quoted above. Although para 6.3.2 appears under the general

heading Design Brief, the requirement for a design life of 75 years is not concerned only

with elements of pure design; it cross refers to the broader requirement for reliable service

during that period. What the employer is entitled to expect under this provision is that at

hand over the tunnel will be designed and built to a standard where it can be expected that

it will provide reliable service for 75 years without requiring major refurbishment or

significant capital expenditure. It is not enough that, according to the preponderance of

expert evidence, the design was suitable for a tunnel expected to last without major

expenditure for that period, or that the tunnel “ought to have lasted for the 75 year period,

at least in the ground conditions then understood to exist and as set out in the relative

report” (adopting your Lordship’s summary of the position in para [255]). Implementation

of that design, having regard to the ground conditions actually encountered, is also directly

relevant. If it is discovered during the two year defects period before issue of the Defects

Certificate that the tunnel did not have a design life of 75 years, then that part of the works is

not in accordance with the Works Information and there is a Defect within the definition of

[371] The pursuers’ argument is simple and attractive. The HRT, as built, did not have a

design life of 75 years. It did not even last six months. It was therefore not in accordance

with the Works Information of which paragraph 6.3.2 forms part. That is sufficient to

[372] The commercial judge made no specific finding on whether there was a Defect

within the first limb of the definition in clause 11.2(15) of the contract. This was because he

proceeded on the basis that Option M “placed an important brake on liability”. With respect

to the commercial judge, this is to confuse two separate points. Option M only operates as a

brake on liability for Defects in the works due to the contractor’s design. Proof that the

contractor used reasonable skill and care does not mean that there is not a Defect; rather it

establishes that although there is a Defect, it is not one for which the contractor is liable. So

the question of whether or not there is a Defect has to be answered separately from the

subsequent question of whether it is one for which the contractor is excused liability because

[373] Fortunately, however, in his conclusions on why the tunnel collapsed (at paragraphs

[147] -[152]), the judge makes findings of fact which clearly establish, at least on balance of

probabilities, the cause of the collapse. At paragraph [147], the judge finds that the tunnel

collapsed “because there was not enough support: poor rock conditions coincided with

insufficient shotcrete and rockbolts”. At paragraph [152] the judge describes the most likely

mechanism of the collapse: weak rock deteriorated and lost its strength when submerged;

the flowing water washed out areas of erodible rock; the erosion progressed and opened up

larger seams; and, ultimately, the tunnel lost stability and collapsed. If, therefore, it is

necessary to make detailed findings as to what it was which caused the tunnel to collapse,

the judge’s explanation of the collapse as set out in those paragraphs provides all the

material needed to justify the conclusion that the problem was insufficient support

(shotcrete and/or rockbolts) in way of areas of poor rock conditions. (Those poor rock

conditions, or weak rock, are often described as “erodible rock”, a term which, as your

Lordship points out (at para [290]), did not require a definition but was generally

understood as a reference to a type of rock which was particularly susceptible to erosion on

contact with water.) But in truth, for the purposes of establishing a limb one Defect under

reference to the required 75 year Design Life (paragraph 6.3.2), in circumstances where the

tunnel collapsed within six months of hand over and without there having been some

intervening the event (after hand over) capable of damaging the integrity of the tunnel, it is

not necessary to do more than point to the fact of the collapse within that time and without

[374] For those reasons, expressed perhaps at undue length, I have come to the conclusion

that the pursuers have made good their case that the cause of the collapse was a Defect

which existed at take over, in that a part of the Works, viz the HRT, was not in accordance

with the Works Information, viz the requirement for a tunnel with a design life of the 75

years. In those circumstances, and subject to Option M as discussed in the following

paragraph, the loss or damage to the works by virtue of the collapse of the tunnel was at the

[375] It still remains to be considered whether the defenders are relieved from liability by

the terms of Option M, which provide that 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. I can, however, deal with this point briefly.

As your Lordship explains at paras [256]-[257] , on the judge’s findings there was no defect

in the design as such. The evidence does not support the contention that the tunnel, as

designed, should not have lasted the required 75 years. What appears to have gone wrong

was in the implementation of that design, probably in the failure to identify (at the rock face

in the tunnel) rock conditions requiring Class III and Class IV support, resulting, as the

judge found, in insufficient support being provided to areas of erodible rock. It is

unnecessary to ask whether that rock identification, or mis-identification, reflects a want of

reasonable skill and care – the judge has found that it did not, and I agree with your

Lordship that we cannot interfere with his findings of fact on this issue. But all that is

irrelevant unless the Defect was a Defect in design. It was not. The Defect was not one of

design but rather implementation of that design. It was that failure in implementation of the

design which resulted in the fact that the tunnel on hand over did not have a design life of

75 years. Option M therefore does not come into play at all. It does not relieve the

[376] I should briefly mention the question of causation. The judge dealt with it at para

[195] . He proceeds on the basis that the pursuers’ complaint was that the defenders had not

installed at least Class III support in the areas of erodible rock. He finds, on the evidence,

that only Class IV would have prevented the collapse. I come back to consider this (below)

in connection with the limb two Defect, but this line of reasoning has no application to the

limb one Defect, viz failure to provide the Works with a tunnel which met the requirement

for a 75 year Design Life. Had it met that requirement, it obviously would not have

collapsed within six months of hand over. No causation point arises in respect of this

category of Defect; or, if it does, it is readily answered in favour of the pursuers.

A limb two defect: part of the works not in accordance with Contractor’s accepted design?

[377] Your Lordship deals with this point at paras [258]-[262]. I entirely agree with your

Lordship that the contractor’s design “accepted” by the project manager was the HRT RSM,

which included Table 11. What happened thereafter in the tunnel was, as your Lordship

describes it, “an exercise in the implementation of that design”. Accordingly, for present

purposes, the question to be asked is whether part of the works (these being works which

were designed by the contractor) were not in accordance with the contractor’s design as set

[378] A copy of Table 11 is helpfully attached as Appendix 3 to your Lordship’s Opinion.

It is not, therefore, necessary to set out its terms in any great detail here. The structure of the

Table is to identify hazards likely to be encountered in the HRT; to assess the severity of the

risk associated with such hazards before mitigation measures are taken; to identify

appropriate “countermeasures”; and to re-assess the risk on the assumption that such

countermeasures have been taken. One of the hazards identified, at the bottom of the Table,

is “erosion of erodible rock during operation”. That risk was assessed with a risk rating

score of 9, representing a risk at the top end of a band (coloured in yellow) described as

“high risk”. The countermeasures identified to deal with that risk were: “application of

shotcrete if not already covered/ ‘protected’ by steel rib support”. On the assumption that

such countermeasures would be taken, the risk was reassessed as “low” (coloured in green).

[379] As your Lordship accepts (at para [259]), it may be assumed from the findings of the

judge that some erodible rock had not been shotcreted. Whether this was because certain

sections of rock were not identified as being erodible when in fact they were, or whether it

was because a decision was taken on inspection in the tunnel that shotcreting was not

necessary at that particular point, is of no consequence in my opinion. I should not be taken

as concurring in your Lordship’s understanding (expressed in that same paragraph) that

rock only became erodible after its initial deterioration. If that were the position, it would

mean that no erodible rock as such was ever exposed in the tunnel during the course of

construction, a proposition which I do not understand to be established either from the

evidence or from the findings made by the judge. The encouragement (whether mandatory

or otherwise) in Table 11 to protect erodible rock during construction by shotcreting or steel

support would be meaningless if erodible rock did not yet exist, or, if it did, was neither

visible nor detectable during the course of construction. It is safe, in my opinion, to proceed

upon the assumption that erodible rock existed and was detectable in the area of the

subsequent collapse and was, for whatever reason, not shotcreted or supported in any other

[380] On this basis, the question is whether the absence of shotcrete or other support in

areas of erodible rock was in accordance with the HRT RSM, and more particularly Table 11,

that being the relevant part of the contractor’s design accepted by the project manager. That

is a question of fact to which, in my opinion, there can be only one answer: No.

[381] In my opinion the question does not ultimately turn on whether there was a

“mandatory requirement to shotcrete every area of erodible rock, however unimportant the

area might have seemed to an experienced geological engineer on site” to use your

Lordship’s formulation in para [260]. That would be the correct test if the question was

whether the defenders were in breach of contract by failing to shotcrete particular sections of

erodible rock within the tunnel. But the question here is a subtly different one, relating not

to breach but to the allocation of risk. All that has to be asked for this purpose is whether

there was a Defect as defined in section 11.2(15); and that turns on the question of whether

the works as carried out were or were not “in accordance with” the accepted design. If they

were not in accordance with that accepted design, because of the absence of shotcreting or

other support in way of erodible rock, then there was a Defect as defined in section 11.2(15)

[382] At para [166] of his Opinion, the judge identified what he described as “formidable

difficulties” with the pursuers’ approach to this question. I shall deal with them in turn

[383] The first point made by the judge was that this provision in Table 11 ran counter to

the overwhelming weight of the other contractual provisions which call for the exercise of

engineering judgement in the tunnel. That seems to me, with respect, to merge the design

stage with that of implementation of the design. The design had been agreed and accepted:

erodible rock was to be shotcreted if not otherwise protected. The implementation stage

required the exercise of engineering judgement in identifying which rock was erodible and

required to be dealt with in this way. The fact that engineering judgement was to be applied

at that implementation stage does not detract at all from the fact that the design itself, as

agreed and accepted, specified what measures were to be taken to minimise the risk from

[384] I should add, since the point was pressed hard in argument by Mr McBrearty QC for

the defenders, that I see little force in the argument that this was but one isolated provision

in a technical document. It was in fact the provision which identified what steps were to be

taken if erodible rock was encountered. But even if it were but one isolated provision, that

would not mean that it would fall to be disregarded. It was part of the accepted design. A

similar argument was raised and rejected in MT Højgaard at para 49.

[385] The judge’s second point was that the term “erodible rock” was vague and required

interpretation at the face. So be it. There would be a need at the face during the course of

construction to identify erodible rock, but this was all part of the process of implementing

the accepted design. Erodible rock, once identified during the course of this process, ought

[386] Thirdly, the judge thought that if the pursuers’ argument was correct then

determination of the “overall excavation class” on the REC sheet should have been a

mechanistic process depending upon the worst feature ticked. This may be correct, but the

REC sheets, both in their design and use, were devised as part of the implementation of the

[387] Fourth, the judge thought that if the argument was correct then the pursuers would

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

That may be right, but so what? They had agreed and, through the project manager,

accepted the design which, in Table 11, required a specified level of protection to be given to

erodible rock. That was something which would have to be paid for. Better to pay for the

appropriate level of support than to risk a collapse within the tunnel.

[388] Fifthly, the judge commented that, if the pursuers were correct, it would not have

been possible to have an unlined tunnel in similar geology – the classification would

inevitably have resulted in a lined tunnel. That is not entirely correct. There might well

have been a requirement for a section of the tunnel to be fully or substantially lined, but this

would only apply to that part of the tunnel passing through the CFZ. There is no reason to

suppose that other stretches of the tunnel, both upstream and downstream, would have had

to be fully lined. They were not in fact fully lined and the issue about “Secondaries”, dealt

with by your Lordship at paras [309]-[310], would seem to suggest that this was not in fact a

[389] For these reasons I am persuaded that the pursuers have made out the case that there

was a Defect under this head. The absence of shotcreting or other protection in areas of

erodible rock meant that that part of the works was not in accordance with the contractor’s

design accepted by the project manager. The Defect was not one of design; rather it was in

implementing the design agreed between the parties and accepted by the project manager.

[390] This leads on, again, to the question of causation. The judge makes a finding, at para

[195] , that Class III support (ie full circumference shotcrete) would not have been enough to

avoid the collapse; only Class IV support (increased shotcrete, together with steel arches)

would have prevented it. On that basis he concludes that the pursuers have failed to

establish that this Defect was causative of the collapse. There is, to my mind, a problem

with this line of reasoning. It appears to proceed on the assumption that all that was

required by the pursuers was Class III support. But this is not correct. In the first sentence

of that paragraph the judge identifies the pursuers’ case as being that the defenders should

have installed “at least class III support” (emphasis added). In an earlier paragraph (para

[165] ) the judge identifies the pursuers’ case as being that the defenders “should have

installed class III or IV support wherever it found erodible rock”. Although the pleadings

are opaque on this point, it appears that the judge did not understand the pursuers’ case to

be limited to the use of Class III. Nor did Table 11 confine itself in this way; it required

shotcreting “if not already covered/ ‘protected’ by steel rib support” (ie Classes III and IV as

appropriate). There is no reason to suppose that if the erodible rock had been treated in this

way the collapse would not have been avoided. There is certainly no finding to this effect

by the judge. In my opinion the causation point should be answered in favour of the

[391] For the reasons set out above, I consider that the pursuers have established their case

on Defects under both limbs. Neither being a Defect in design, Option M has no application.

It follows that in terms of clause 80.1 and 81.1 the collapse was at the defenders’ risk and

[392] At paras [317]-[320] your Lordship in the chair in effect adopts the reasoning of the

judge on the joint insurance issue and rejects the defenders’ cross appeal on this point.

Although I concur in rejecting the cross appeal on this point, I differ from Your Lordship as

to the reasons for doing so; and, since the point is of some importance, I should explain my

[393] As your Lordship notes in para [317], the defenders were obliged in terms of clause

84 to provide insurance in the joint names of the parties in respect of (amongst other things)

loss of or damage to the works caused by events which were at their risk prior to the issue of

the defects certificate. The defects certificate is issued at the defects date (104 weeks after

completion of the works) or at the end of the last defects correction period, whichever is

later (clause 43). Accordingly, damage to the tunnel as a result of the collapse, which

occurred within six months or so after completion, would fall within this category,

provided, of course, that the cause of the collapse was at the defender’s risk. The allocation

of risk is dealt with in clauses 80.1 and 81.1. So far as material to the present issue, once the

works had been taken over by the pursuers (which they had been), they were at the

pursuers’ risk unless, among other things, the loss or damage occurred before the issue of

the defects certificate and was due to a defect existing at take over; in which case they were

at the defenders’ risk. In respect of loss of or damage to the works (ie the tunnel), the

[394] Clause 85 required the insurance policy taken out by the defenders to include a

waiver of the insurer’s rights of subrogation against the insured’s employees. Although this

provision did not itself stipulate for the policy to contain a waiver of subrogation in claims

against the insured themselves, the policy in fact taken out by the defenders, and

presumably accepted by the Project Manager on behalf of the pursuers (clause 85.1), did in

fact contain such a waiver of subrogation rights against the insured.

[395] In terms of clause 83.1, each party agreed to indemnify the other against claims,

proceedings, compensation and costs due to an event which was at that party’s risk.

[396] Since the insurance required to be taken out under clause 84 was only to cover loss

and damage due to events which were at the defenders’ risk, the defenders’ argument has to

assume, contrary to their primary case, that the tunnel collapse was indeed due to an event

which was at their risk. On that hypothesis, their argument is straightforward. Where

parties have agreed to enter into a joint insurance in respect of risks of loss and damage to

the works, in the event of such loss or damage they must look to the insurance policy for

redress and cannot bring claims against one another in respect of that damage. They rely on

a long line of English cases, culminating in the judgments in the House of Lords and the

United Kingdom Supreme Court in CRS and the Ocean Victory respectively, to the effect that

the existence of a joint insurance provision normally gives rise to an implied term to that

[397] The matter was argued at debate on the basis of the parties’ pleaded cases. It

appears at that stage (though no such distinction was drawn in the argument before this

court) to have been accepted on behalf of the defenders that the joint insurance argument

did not respond to the claim brought by the pursuers under the Defects provisions in

clauses 43 and 46.4, and (possibly) did not respond to the claim under clause 82.1: see paras

[44] -[48] of the judge’s Opinion of 14 July 2015 [2015] CSOH 92. But it was argued that the

joint insurance did preclude a claim by the pursuers for an indemnity under clause 83.1.

The judge rejected that argument. He held, in short, that an implied term of the sort

contended for would render clause 83.1 redundant (paras [81]-[82]); he pointed to the

absence from the contract of a requirement for a waiver of subrogation rights in respect of

the parties themselves (para [83]); and he attached importance to the terms of clause 85.4

which provided that amounts not recovered from the insurer should be borne by the parties

according to where the risk lay (para [84]). In summary, he held (para [87]) that an implied

term of the sort contended for by the defenders was not necessary for the purposes of

business efficacy and would do violence to the language selected by the parties; there was

no irrebuttable presumption that parties had no liability to one another simply because a

joint names policy is in place and to hold otherwise would be “to merge the law of insurance

[398] As the judge recognised at para [79], the thrust of the authorities is strongly in favour

of joint names insurance displacing contractual liability. That is confirmed by the judgments

of the Supreme Court in the Ocean Victory, which was decided after the judge reached his

decision in this case. Although none of the relevant cases are Scottish, and the observations

of the various judges on this point are all obiter, I agree with your Lordship in the chair that

the reasoning on this point is compelling and of high authority and should be followed in

[399] Although, as already observed, the judge noted that the thrust of the authorities is

strongly in favour of joint names insurance displacing contractual liability, it is nonetheless

worth pausing to note the reasons for this and the strength of that presumption. The Ocean

Victory involved an appeal by hull insurers (as assignees of claims by owners and demise

charterers) against a finding by the Court of Appeal that time charterers were not in breach

of a safe port warranty in the charter party. That appeal was dismissed unanimously by the

Supreme Court. However the members of the court also expressed views on the effect of a

provision in the demise charter for joint insurance. On this point they were divided.

[400] Lord Toulson (with whom Lords Mance and Hodge agreed) categorised the question

of whether the contractual scheme precluded a claim for the insured loss as one of

“The question in each case is whether the parties are to be taken to have intended to

create an insurance fund which would be the sole avenue for making good the

relevant loss or damage, or whether the existence of the fund co-exists with an

independent right of action for breach of a term of the contract which has caused that

He relied (at para 140) upon the decision in CRS for the proposition that:

“… it cannot have been the parties’ intention that parties who were jointly insured

under a contractor’s all risks policy could make claims against one another in respect

of damage covered by the insurance, or that the insurers could make a subrogated

claim … and that the court would if necessary hold that there was an implied term to

Lord Toulson approved the reasoning of Mr Recorder Jackson QC in Hopewell Project

Management v Ewbank Preece [1998] 1 Lloyd’s Rep 448 at 458, noting that such an implied

term presupposed that the party relying on it had not prevented recovery under the policy.

The insurance arrangements provided not only a fund but also a term for the avoidance of

[401] Lord Mance made additional observations (at para 114) to the effect that the

principle was best viewed as resting on the “natural interpretation of or implication from the

contractual arrangements giving rise to such co-insurance”, citing Lord Bingham in CRS at

“… anyone contemplated that the co-insurance to be arranged could give rise to

successive payments of the same sum to different parties, with the second of such

He concluded (at para 122) that the reason why there was no claim against the party in

default was not that such a claim existed but was discharged by receipt of payment from the

“… under a co-insurance scheme like the present, it is understood implicitly that

[402] Lord Sumption, with whom Lord Clarke agreed on this point, took the same view on

this general principle. He explained (at para 98) that in the case of joint insurance, or

insurance taken out pursuant to a term of the contract for the benefit of one of the parties

thereto, the collateral benefits rule (often expressed in the maxim that insurance is res inter

alios acta or, loosely translated, none of his business) had no application since the insurance

“manifestly is the wrongdoer’s business”. He expressed the general principle in this way (at

“It is well established, and common ground between the present parties, that where

it is agreed that the insurance shall inure to the benefit of both parties to the contract,

they cannot claim against each other in respect of an insured loss. Co-insurance is

the paradigm case. … What is less clear is its juridical basis. … The better view,

which was endorsed by the House of Lords in [CRS], paras 61-65 (Lord Hope of

Craighead), is that it is an implied term of the contract of insurance and/or of the

underlying contract between the co-insureds pursuant to which their interests were

insured. The implication is necessary because if the co-insureds are both insured

against the relevant loss, the possibility of claims between them is financially

irrelevant. It would be absurd for the insurer to bring a subrogated claim against a

co-insured whom he would be liable to indemnify against having to meet it. …”

Thus far there is no difference between what he says and the views of the majority.

[403] What is important for present purposes, and this is the reason for citing these

passages, is to note the strength of the presumption. Words such as “inconceivable” and

“absurd” suggest the need for powerful contra-indicators if such a term excluding

[404] I should mention at this point that Lord Sumption went on (in para 100) to raise a

different question directed to identifying the basis for the rule. He said this:

“When we say that one co-insured cannot claim damages against another for an

insured loss, is that because the liability to pay damages is excluded by the terms of

the contract, or is it because as between the co-insureds the insurer’s payment makes

good any loss and thereby satisfies any liability to pay damages?”

He went on to give an illustration of why that point might be significant. He concluded that

the correct analysis would depend upon the particular terms of the particular contract and

that the answer would not necessarily be the same in every case. In the course of argument

in this case the court was told that there was here a significant shortfall between the amount

of the insurance proceeds payable under the policy and the actual cost of the repairs to the

tunnel. This case may therefore provide an illustration of a situation in which the question

raised by Lord Sumption might be critical. It may also illustrate the need to consider what

other qualifications there might be to the general principle adumbrated in the decided cases:

for example, what is the position if, through the fault or neglect of one or other party, the

insurance is not taken out in the terms agreed in the contract; or if insurers are entitled to

avoid liability; or if the agreed insurance proves inadequate to cover the amount of the loss?

However, there are no pleadings on this matter, nor was any evidence led on it; and neither

side sought to develop the argument in any detail. In those circumstances it would not be

appropriate for this court to enter into a detailed discussion of this point and it is, in any

event, unnecessary to do so since the joint insurance point can be disposed of on other

[405] For present purposes, what is important is the recognition that the general principle

exists and is based upon the proper construction of the contractual arrangements between

the parties or an implied term thereof; the difference is one of expression only and is

immaterial. It should be noted that the principle does not depend, as was thought at one

time, upon any consideration of circuity of action; and, since it does not depend upon

considerations of circuity, it must in my view now be recognised that the question whether

the contractual documents or the policy of insurance contains a waiver of rights of

[406] The judge accepted the existence of the principle that joint names insurance generally

displaced contractual liability but recognised, correctly, that whether it applied in the

present case depended upon the terms of this particular contract. He considered that it did

not apply for the reasons summarised above. I differ from the judge on this point.

[407] The first of the reasons given by the judge was that, so he thought, the implication of

such a term ran counter to clause 83.1 and would effectively render clause 83.1 redundant. I

find this difficult to understand, since clause 83.1 extends well beyond loss and damage of

the types covered by the insurance clause (clause 84); provides for mutual indemnities

whereas the insurance clause is concerned only with loss and damage occurring as a result

of events at the defenders’ risk; and is not time restricted to the same extent as the insurance

[408] Secondly, the judge relied upon the absence in clause 85 of any waiver of

subrogation rights as against the other party to the contract; but as he himself had made

clear (in paras [20]-[21]), the policy itself, which was entered into as a matter of agreement

between the parties, did contain such a waiver of subrogation rights. And further, as I have

sought to explain, the existence or otherwise of a waiver of subrogation rights is irrelevant to

the application of the principle now that it is recognised that it arises as an implied term of

the contract rather than out of the need to avoid circuity of action.

[409] As to the judge’s third point, namely that under clause 85.4 it is provided that the

amounts not recovered from insurers shall be borne by the parties in accordance with the

allocation of risk under the contract, there is scope for the application of that provision in

respect of items such as deductibles; it is not a pointer against implication of the joint

insurance term. It is possible too that that clause would be apt to cover a situation such as

that contemplated by Lord Sumption in the passage referred to, where the insurance does

not for whatever reason answer to the full amount of the claim, but this does not detract

from the general presumption as to the existence of the implied term; and, in any event,

there are no averments (and there has not been any evidence) as to amounts not recovered

[410] In my opinion, none of these three factors relied on by the judge, taken individually

or together, point to an intention to exclude the implied term to the effect that joint names

[411] Accordingly, there is in my view force in the defenders’ argument that had the claim

been brought simply on the basis of a claim for damages for breach of contract leading to the

collapse of the tunnel and/or on the basis of a claim for an indemnity pursuant to clause

83.1, it could have been met by the defenders by relying upon the implied term identified in

the cases arising from the agreement between the parties to effect insurance in joint names.

[412] However, this does not assist the defenders in the present case. The claim advanced

under clause 43.1 and 46.4 is in respect of the failure by the defenders to correct Defects

and/or, in default of so doing, to pay the pursuers the costs of carrying out the work

themselves. Similarly, the claim under clause 82.1 is in respect of a failure by the defenders

to replace and repair loss and damage to the works. These are claims for failure to comply

with specific provisions requiring the defenders to carry out repair or reinstatement works

similar to those discussed by Lords Hope and Rodger on a hypothetical basis in CRS (at

paras 49 and 71 respectively). The implied term arising from the provision for joint

insurance would not prevent the pursuers succeeding in these claims for the simple reason

that the joint insurance required to be taken out pursuant to clause 84 was insurance against

loss or damage to the works and did not cover breach of contract by the defenders in failing

to carry out specific obligations under the contract. Had the defenders in fact corrected the

defects in accordance with clause 43.1 or, as the judge correctly found that they ought to

have done, replaced and repaired the damage to the works in accordance with clause 82.1,

or had they paid the pursuers the cost of carrying out such work in default of doing it

themselves, they might have sought to recover their costs of so doing from the pursuers

pursuant to clause 83.1. But that would have been on the basis that, as they would have had

to contend, the relevant risk lay with the pursuers – and that would not have been covered

by the joint insurance which only answers to events at the defenders’ risk.

[413] For these reasons I consider that the joint insurance point does not assist the

defenders and, albeit for different reasons, I too would dismiss the cross-appeal on this

[414] In my opinion the reclaiming motion should be allowed on what I have called the

[415] So far as concerns the consequences of this, the judge dealt with the financial

consequences in his Note of 31 January 2017. He indicated that if, contrary to his own view,

the defenders were liable for the collapse of the tunnel because that collapse was a

contractor’s risk event, then he would have awarded the pursuers the sum of

£107,617,830.94 in respect of the cost of the recovery project: see paras 6-8 of that Note. I see

no reason to depart from this assessment. I would award interest on this sum on the same

basis as the judge, namely at the rate of 4% per annum from the date of citation until the date

of the decree pronounced by this court, and thereafter at the rate of 8% until payment. That

sum is in addition to the £1,000,000 awarded by the judge as low availability damages and

the sums of €388,720.27 and £32,357.98 paid by the pursuers under the second adjudicator’s

award which the judge held to be recoverable by them: see paras (1) and (2) of the

interlocutor dated 2 February 2017. On those sums I would award interest at the rate of 4%

per annum from the date of citation until 2 February, being the date of the decree

pronounced by the judge, and thereafter at the rate of 8% until payment.

[416] I would reserve all questions of expenses. Although the appeal on the expenses

point (as a discrete point) is to be refused, if the appeal on the Defect Issue succeeds then

expenses both in this court and before the judge will be at large.

Scottish Court of Session Decisions