Canadian National Railway Co. et al. v. Royal and Sun Alliance Insurance Co. of Canada et al.

[Indexed as: Canadian National Railway Co. v. Royal and Sun Alliance Insurance Co. of Canada]

85 O.R. (3d) 186

Court of Appeal for Ontario,

Rosenberg, Cronk and Lang JJ.A.

March 26, 2007

Insurance -- Exclusion clauses -- Builders risk policy excluding from coverage cost of making good "faulty or improper design" -- Tunnel boring machine breaking down and causing delay to project in which it was being used -- Breakdown caused by excess differential deflection between components of machine which machine was not capable of withstanding -- In determining whether exclusion applied, trial judge properly applying standard of foreseeability rather than tort standard of reasonable foreseeability -- Trial judge erring in finding that excess differential deflection was not foreseeable risk so that faulty or improper design exclusion did not apply -- Trial judge not erring in finding that "inherent vice" exclusion did not apply.

The plaintiffs retained L Inc. to construct a custom-designed tunnel boring machine (the "TBM") for use in the construction of a railway tunnel. The TBM broke down, resulting in damage to the TBM itself and a 229-day delay in the opening of the tunnel. The plaintiffs held a builders risk insurance policy issued by the defendants that insured against all risks of direct physical loss or damage to all real and personal property of every kind and quality, including but not limited to the TBM. When the plaintiffs sought indemnity under the policy for their losses, the defendants denied coverage, relying on exclusions in the policy that exempted from coverage "the cost of making good . . . faulty or improper design" and "inherent vice". The trial judge held that the cause of the failure of the TBM was excess differential deflection between components of the TBM which the TBM was not capable of withstanding; that this damaged the TBM's sealing system, thereby allowing contaminants to access the TBM's main bearing; that the excess differential deflection that caused the failure of the TBM was not foreseeable; that the coverage exclusions invoked by the defendants did not apply; and that a policy exception to the faulty or improper design exclusion for "resultant loss or damage" did not provide coverage in this case. If, contrary to his conclusion, the faulty or improper design exclusion applied, he held that this exclusion extended to all loss or damage caused by the faulty or improper design of the TBM. Finally, he held that no physical loss or damage to the tunnel itself was occasioned by the failure of the TBM within the meaning of the coverage provisions of the policy. Judgment was granted in favour of the plaintiffs. The defendants appealed, and the plaintiffs cross-appealed.

Held, the appeal should be allowed; the cross-appeal should be dismissed.

Per Rosenberg and Cronk JJ.A.: The trial judge erred by holding that the faulty or improper design exclusion did not apply. The central point at issue on the appeal was the trial judge's finding that excess differential deflection was not a foreseeable risk, as his conclusion that the faulty or improper design exclusion could not be relied on by the defendants rested on his foreseeability findings. The trial judge's foreseeability findings were erroneous for several reasons. First, he focused on the expert engineering evidence about foreseeability and failed to consider the evidence of the principal of L Inc. concerning the foreseeability of risk. That evidence established that L Inc. knew when the TBM was designed that [page187] differential deflection posed a serious potential danger to the TBM's sealing system. More importantly, the extent of this potential danger was actually investigated. Second, his foreseeability findings were apparently based, at least in part, on a misapprehension of the evidence. Third, the trial judge erred by failing to distinguish between the foreseeability of the type of risk of failure that occurred here and the foreseeability of the mechanism by which such risk might materialize. That error went to the heart of his foreseeability analysis. The standard of foreseeability endorsed by the trial judge required that all known and foreseeable risks of failure be addressed in the design of the TBM. This did not occur. Instead, contrary to the findings of the trial judge, the design of the TBM was inadequate to meet a known risk of possible failure of the TBM. The design of the TBM was faulty, and the faulty or improper design exclusion applied.

The trial judge did not err in applying the standard of foreseeability to the construction of the faulty or improper design exclusion rather than the tort standard of reasonable foreseeability.

The trial judge did not err in finding that the resulting loss or damage exception to the faulty or improper design exclusion did not provide coverage.

Per Lang J.A. (dissenting): The governing standard only requires that all foreseeable risks be taken into account. The design is not required to succeed in accommodating those risks. The trial judge properly applied the correct standard and found that L Inc. took into account all foreseeable risks. On the facts of this case, L Inc. recognized the risk of excess differential deflection and undertook all precautions to exclude the risk in the design and manufacture of the TBM. To interpret "faulty design" as meaning any design defect, whether that defect was sufficiently taken into account or not, would run contrary to the reasonable expectations of the parties.

APPEAL AND CROSS-APPEAL from the judgment of Ground J., 2004 33029 (ON SC), [2004] O.J. No. 4086, [2004] O.T.C. 851 (S.C.J.), in an action against insurers.

Cases referred to Algonquin Power (Long Sault) Partnership v. Chubb Insurance Co. of Canada, [2003] O.J. No. 2019, [2003] O.T.C. 446 (S.C.J.); Foundation Co. of Canada Ltd. v. American Home Assurance Co., [1997] O.J. No. 2332, 71 A.C.W.S. (3d) 957 (C.A.), affg (1995), 1995 7102 (ON SC), 25 O.R. (3d) 36, [1995] O.J. No. 2164 (Gen. Div.); Hartford Fire Insurance Co. v. Benson & Hedges (Canada) Ltd., 1978 37 (SCC), [1978] 2 S.C.R. 1088, [1978] S.C.J. No. 53, 85 D.L.R. (3d) 467, consd British Columbia v. Royal Insurance Co. of Canada, 1991 5732 (BC CA), [1991] B.C.J. No. 2959, 60 B.C.L.R. (2d) 109, [1992] I.L.R. Â1-2816 (C.A.) [Leave to appeal to S.C.C. refused [1991] S.C.C.A. No. 506], distd Other cases referred to B.C. Rail Ltd. v. American Home Assurance Co., 1991 5713 (BC CA), [1991] B.C.J. No. 697, 54 B.C.L.R. (2d) 228, 79 D.L.R. (4th) 729, [1992] I.L.R. Â1-2789 (C.A.); Boucher v. Public Accountants Council for the Province of Ontario (2004), 2004 14579 (ON CA), 71 O.R. (3d) 291, [2004] O.J. No. 2634, 188 O.A.C. 201, 48 C.P.C. (5th) 56 (C.A.); C.C.R. Fishing Ltd. v. British Reserve Insurance Co., 1990 145 (SCC), [1990] 1 S.C.R. 814, [1990] S.C.J. No. 34, 45 B.C.L.R. (2d) 145, 69 D.L.R. (4th) 112, 109 N.R. 1, [1990] 3 W.W.R. 501, [1990] I.L.R. Â1-2582 (sub nom. C.C.R. Fishing Ltd. v. Tomenson Inc.); Celanese Canada Inc. v. Canadian National Railway Co., 2005 8663 (ON CA), [2005] O.J. No. 1122, 196 O.A.C. 60 (C.A.) [Leave to appeal to S.C.C. refused [2005] S.C.C.A. No. 245]; Collavino Inc. v. Employers Mutual Liability Insurance Co. of Wisconsin, [1985] O.J. No. 227, 14 C.C.L.I. xli (C.A.), affg 1984 5963 (ON SC), [1984] O.J. No. 1011, 7 C.L.R. 165 (H.C.J.); Consolidated Bathurst Export Ltd. v. Mutual Boiler and Machinery Insurance Co., 1979 10 (SCC), [1980] 1 S.C.R. 888, [1979] S.C.J. No. 133, 112 D.L.R. (3d) 49, 32 N.R. 488, [1980] I.L.R. Â1-1176; [page188] Dawson Creek (City) v. Zurich Insurance Co. (2000), 2000 BCCA 158, [2000] B.C.J. No. 463, 75 B.C.L.R. (3d) 131, [2000] 5 W.W.R. 311, [2000] I.L.R. Â1-3825 (C.A.) [Leave to appeal to S.C.C. refused [2000] S.C.C.A. No. 217]; Employers Casualty Co. v. Holm, 393 S.W. 2d 363 (Tex. Civ. App., 1965); Graham v. Rourke (1990), 1990 7005 (ON CA), 75 O.R. (2d) 622, [1990] O.J. No. 2314, 40 O.A.C. 301, 74 D.L.R. (4th) 1 (C.A.); Keljanovic Estate v. Sanseverino, 2000 5711 (ON CA), [2000] O.J. No. 1364, 186 D.L.R. (4th) 481, 9 E.T.R. (2d) 32 (C.A.) [Leave to appeal to S.C.C. refused [2000] S.C.C.A. No. 300]; Queensland Government Railways and Electric Power Transmission Pty., Ltd. v. Manufacturers' Mutual Insurance, Ltd., [1969] 1 Lloyd's L.R. 214 (H.C.A.); Schaible Electric Ltd. v. Melloul-Blamey Construction Inc., 2005 26706 (ON CA), [2005] O.J. No. 3226, 201 O.A.C. 71 (C.A.); Soya G.m.b.H. Mainz Kommanditgesellschaft v. White, [1983] 1 Lloyd's L.R. 122 (H.L.); Walker v. Ritchie, [2006] 2 S.C.R. 428, [2006] S.C.J. No. 45, 2006 SCC 45, varg 2005 13776 (ON CA), [2005] O.J. No. 1600, 197 O.A.C. 81 (C.A.); Waxman v. Waxman, 2004 39040 (ON CA), [2004] O.J. No. 1765, 186 O.A.C. 201, 44 B.L.R. (3d) 165 C.A.), supp. reasons 2004 31900 (ON CA), [2004] O.J. No. 6013, 6 B.L.R.(4th) 167 (C.A.) [Leave to appeal to S.C.C. refused [2004] S.C.C.A. No. 291]; Willowbrook Homes (1964) Ltd. v. Simcoe and Erie General Insurance Co., 1980 ABCA 139, [1980] A.J. No. 855, 22 A.R. 95 (C.A.) Statutes referred to Courts of Justice Act, R.S.O. 1990, c. C.43, s. 130(1)(b) Authorities referred to Brown, C., Introduction to Canadian Insurance Law, 2nd ed. (Markham, Ont.: LexisNexis Butterworths, 2006)

Earl A. Cherniak, Q.C., Kirk F. Stevens and David E. Liblong, for appellants (respondents by cross-appeal). Richard H. Shaban and Sharon C. Vogel, for respondents (appellants by cross-appeal).

ROSENBERG AND CRONK JJ.A.: --

I. Introduction

[1] In the early 1990s, the respondents, Canadian National Railway Company and allied corporations (the "CN companies"), undertook the construction of a new railway tunnel under the St. Clair River between Sarnia, Ontario and Port Huron, Michigan. Lovat Tunnel Equipment Inc. ("Lovat"), an experienced tunnel equipment manufacturer, was selected to construct a custom-designed tunnel boring machine (the "TBM") for the project.

[2] The TBM began boring the tunnel in November 1993. Approximately two months later, it broke down, resulting in damage to the TBM itself and a 229-day delay in the opening of the new tunnel. [page189]

[3] The CN companies held a builders risk insurance policy issued by the appellant insurers that insured them against "ALL RISKS of direct physical loss or damage . . . to all real and personal property of every kind and quality including but not limited to the [TBM]" (the "Policy"). When the CN companies sought indemnity under the Policy for their losses, the insurers denied coverage, relying on exclusions in the Policy that exempted from coverage "the cost of making good . . . faulty or improper design" and "inherent vice". The CN companies sued the insurers for recovery under the Policy.

[4] A lengthy trial ensued before Ground J. of the Superior Court of Justice. As relevant to these proceedings, he held that: (i) the cause of the failure of the TBM was "excess differential deflection between components of the TBM which the TBM was not capable of withstanding"; [^1] (ii) this damaged the TBM's sealing system, thereby allowing contaminants to access the TBM's main bearing; (iii) the excess differential deflection that caused the failure of the TBM was not foreseeable; (iv) the coverage exclusions invoked by the insurers did not apply; (v) a Policy exception to the faulty or improper design exclusion for "resultant loss or damage" did not provide coverage in this case; (vi) if, contrary to his conclusion, the faulty or improper design exclusion applied, this exclusion extended to all loss or damage caused by the faulty or improper design of the TBM; and (vii) no physical loss or damage to the tunnel itself was occasioned by the failure of the TBM, within the meaning of the coverage provisions of the Policy.

[5] By judgment dated October 7, 2004, the trial judge held the insurers liable under the Policy for damages in the sum of $20,966,947. In so doing, he reduced the damages claimed by the CN companies by about $440,000 on account of the benefit realized by the CN companies on the settlement of an action commenced by them against Lovat. The trial judge also awarded the CN companies prejudgment interest in the amount of $8,615,691.90 based on an average prejudgment interest rate calculated from August 23, 1995, the date that the insurers denied coverage under the Policy. Finally, the trial judge awarded the CN companies their costs of the action, fixed in the amount of $1,150,837.35. [page190]

[6] The insurers appeal the liability finding against them. As ultimately argued before this court, they advanced two grounds of appeal. First, they submitted that the trial judge erred by holding that the excess differential deflection that damaged the TBM's sealing system was not a foreseeable risk and that design error was not established at trial, with the result that the faulty or improper design exclusion did not apply. In the alternative, the insurers asserted that the trial judge erred in his interpretation of the inherent vice exclusion by holding that it had not been made out by the insurers.

[7] In the event that the insurers' appeal is successful, the CN companies cross-appeal from: (i) the trial judge's application of a "foreseeability standard", rather than a "standard of reasonable foreseeability", to the construction of the faulty or improper design exclusion; (ii) the trial judge's interpretation of the scope of the faulty or improper design exclusion; (iii) the trial judge's interpretation of the resultant loss or damage exception to the faulty or improper design exclusion; and (iv) the trial judge's finding that there was no physical loss or damage to the tunnel occasioned by the failure of the TBM, within the meaning of the coverage provisions of the Policy. The CN companies also cross-appeal, on various grounds, from the quantum of damages and costs awarded by the trial judge, and from his calculation of prejudgment interest.

[8] For the reasons that follow, we would allow the appeal. In our view, the trial judge erred by holding that the faulty or improper design exclusion did not apply. We see no error, however, in his conclusion that the inherent vice exclusion was not engaged in this case.

[9] We would dismiss the cross-appeal. We see no error in the trial judge's approach to the construction of the faulty or improper design exclusion. We agree with the trial judge that the resulting loss or damage exception to the faulty or improper design exclusion does not provide coverage, and that this exclusion extends to all loss or damage caused by the faulty or improper design of the TBM, including the costs incurred in inspecting, repairing and modifying the TBM. Nor do we see any error in his holding that there was no direct physical loss or damage to the tunnel arising from the failure of the TBM. Finally, given the view that we take of the proper disposition of these proceedings, the remaining grounds of cross-appeal raised by the CN companies -- concerning the quantum of damages and costs awarded to them and the trial judge's calculation of prejudgment interest -- are moot. [page191]

II. Policy Provisions

[10] The appellants are subscribing insurers to the Policy, under which St. Clair Tunnel Construction Company [^2] and all affiliated, associated or allied corporations are insured in accordance with the following insuring agreement:

1. INSURING AGREEMENT:

This Policy, subject to the limitations, exclusions, terms and conditions hereinafter mentioned, insures, in respect to occurrences happening during the period of this Policy against ALL RISKS of direct physical loss or damage, including general average and salvage charges to:

(a) All real and personal property of every kind and quality including but not limited to the [TBM] . . . ;

(Emphasis added)

[11] The issues in these proceedings primarily concern the interpretation of the following two exclusions in Section I of the Policy:

3. EXCLUSIONS: This Policy does not insure: . . . . .

(d) the cost of making good . . . . .

(iii.) faulty or improper design

provided, however, to the extent otherwise insured and not otherwise excluded under this Policy, resultant loss or damage under any Section of this Policy shall be insured . . . ;

(e) wear, tear, inherent vice, normal upkeep and normal making good; but this exclusion shall not apply to resulting loss not otherwise excluded by this Policy;

(Emphasis added)

[12] Section IV of the Policy, entitled "General Conditions", contains a "sue and labour" clause. In material part, it provides:

16. SUE AND LABOUR:

In case of loss or damage, or threatened loss or damage, it shall be lawful and necessary for the Insured ... to sue, labour and travel for in and about the defence, safeguard and recovery of the property insured hereunder, or any part thereof, without prejudice to this insurance; the Insurer(s) will contribute pro rata towards any [page192] reasonable and proper expenses in connection with the foregoing, according to the respective interests of the parties.

III. Facts

(1) TBM's Sealing System

[13] The TBM was a huge and complex machine. It was the largest TBM ever built by Lovat, being approximately 25 per cent larger than any of the Lovat TBMs previously manufactured. Indeed, at the time, it was the largest TBM of its kind in the world.

[14] The TBM was cylindrical in shape, with a diameter of approximately 32 feet and a length of 278 feet. It had a circular rotating cuttinghead at the front of its forward shell, to which various excavation tools could be attached. The cuttinghead was driven by hydraulic thrust generated by a large main bearing located in a chamber in a static part of the TBM's forward shell. When the cuttinghead was operational, excavated materials fell through openings in the cuttinghead into a hollow chamber (the plenum), from which they were eventually removed from the tunnel by means of a conveyor and auger system.

[15] The TBM was a soft "earth pressure balance" [^3] machine, which was intended to maintain "equality" between the pressure of the earth on the cuttinghead, on the one hand, and the pressure of the cuttinghead on the earth, on the other hand. Maintenance of this equality prevented overburden soils on the top of the TBM from settling or heaving.

[16] The TBM's main bearing was critical to the functioning of the entire machine. To shield it from damage by contamination from soil and other materials, the TBM was designed with a unique and extensive sealing system, the sole purpose of which was to prevent soil and other materials from accessing, and thereby contaminating, the main bearing. This consisted of a series of 26 separate seals, most of which were affixed to a seal ring located between the cuttinghead and the "bulkhead" at the face of the static chamber of the TBM's forward shell. Some of the seals were also attached to the cuttinghead and the bulkhead, or were located in the main bearing itself. While the TBM was in use, pressurized grease was injected into the sealing system to prevent the ingress of soil and other contaminants. [page193]

[17] For the TBM's seals to be effective, a "seal gap" of no less than three millimetres and no more than six millimetres had to be maintained between the rotating cuttinghead and the static bulkhead of the TBM. This was imperative because if the seal gap was too large, grease would pour out of the machine or materials would flow in. Conversely, if the seal gap fell below approximately three millimetres, the seals would be rendered ineffective and possibly crushed, allowing soils or other contaminants to bypass the seals and access the main bearing of the TBM. Thus, the maintenance of the seal gap was vital to the proper functioning of the sealing system, which, in turn, was essential to protect the TBM's sensitive main bearing.

[18] The design of the sealing system was described by various witnesses at trial as conservative, consisting of a system of "redundancy". Among other matters, this meant that each seal operated independently of the others. For the purpose of protecting the main bearing, the sealing system was the TBM's "Maginot Line": to penetrate the vulnerable main bearing, dirt or contaminants had to get past all the seals.

[19] The critical need to prevent contamination of the TBM's main bearing was recognized from the outset of Lovat's engagement. For example, Lovat's construction contract with CN included the following specification:

The design shall be such that the head seals and bearings [sic] are not subject to damage from excessive wear or contamination from soil entering the bearing.

[20] Lovat had previously designed and built 124 TBMs. This TBM, however, was distinctive in several respects. First, each TBM, including each TBM cuttinghead system, was unique and designed for a specific application. In this case, the cuttinghead design configuration for the TBM had not been manufactured or used by Lovat before. Second, although the same type of seals that were used in this TBM had been employed in other Lovat TBMs, their configuration varied from machine to machine. The sealing system on this TBM, including the geometry of the seal cavity, was unique. Third, although component seals of the type installed as part of this TBM's sealing system had functioned successfully in other Lovat TBMs (including at pressures greater than those encountered in the St. Clair River Tunnelling Project), the whole of the sealing system had not previously been tested by Lovat in a TBM of this type and size. Finally, the enormous size of this TBM distinguished it from predecessor TBMs. As CN's expert engineering witness at trial, Dr. Leslie G. Hampson, put it: "The St. Clair TBM was breaking new technical ground being the biggest machine of its type." [page194]

(2) Design process

[21] The process utilized to design and construct the TBM was elaborate and sophisticated, involving many of the world's foremost TBM and tunnelling experts.

[22] The design process included the formation of a technical committee, composed of expert tunnelling contractors and consultants, to advise the CN companies on the conceptual design parameters for the TBM, among other matters. This committee's work was guided by a steering committee, which furnished general guidance on the construction of the TBM. Mr. Rick Lovat, a mechanical engineer and then the vice-president of Lovat who was responsible for overseeing the conceptual design of Lovat's products, participated in the meetings of the technical committee. Although this committee was dissolved before tenders were called for the tunnel contract, its members developed a list of the key components of the TBM.

[23] In addition, Hatch Mott MacDonald ("Hatch"), an engineering firm, was engaged to provide engineering managerial services for the tunnel project. As well, a technical review committee was struck to monitor, review and advise the CN companies on the project as a whole. This review committee, which received presentations by a team of tunnelling experts, met frequently prior to the commencement of the tunnelling and at least twice after the failure of the TBM.

[24] Importantly, both the technical and technical review committees were concerned chiefly with conceptual, rather than detailed, design issues regarding the TBM. Perhaps for this reason, as we mention later in these reasons, these committees were not provided with certain test results and other data available to Lovat regarding the design and the anticipated performance of the TBM.

[25] Instead, once Lovat won the tender for the tunnel project, it was responsible for the detailed design and engineering of the TBM, and for its structural integrity. Rick Lovat testified at trial that Lovat was also responsible for ensuring that the structural components of the TBM did not "deflect or deform in any unwanted way" when the machine was operational.

[26] After Lovat was engaged, it retained Wardrop Engineering Inc. ("Wardrop") to conduct finite element analyses on key components of the TBM, including the outer shell, the cuttinghead, the cuttinghead flange, the bearing assembly (including the seal ring) and the fixed bulkhead. These computer modelling analyses were intended to simulate and predict how the tested elements of the TBM would react under anticipated load, stress and strain conditions. In particular, one of the purposes of the analyses was [page195] to determine whether, and to what extent, critical components of the TBM might deflect under assumed conditions, and the impact of such deflection should it occur.

[27] The modelling inputs, including the anticipated design loads for the TBM, were either provided to Wardrop by Lovat directly or by geotechnical engineers at Lovat's request. In its final report, Wardrop concluded that stresses from the anticipated design loads for the TBM were within acceptable yield criteria, with the result that the TBM was "capable of resisting the loads provided by Lovat which account for the anticipated operation [sic] conditions" of the TBM. According to Rick Lovat, Lovat understood from this report that the modelled components of the TBM might deflect under various loads but that they would maintain their elasticity, that is, they would revert to their original shape and not "break or deform".

[28] The Wardrop analyses revealed the predicted absolute deflection of each of the modelled TBM components. They did not address, however, the anticipated relative deflection of all the components on an "as assembled" basis. Wardrop cautioned Lovat that the determination of "a more representative deflection" of the bearing plate assembly, cuttinghead and forward shell of the TBM would require an additional detailed finite element analysis. Wardrop also stated:

We anticipate the results of the more detailed analysis would reflect a reduced deflection of the bearing plate area because of the increased stiffness from the thrust bearing, cutting- head, and bearing plate interface.

(Emphasis added)

[29] No further finite element analysis was commissioned by Lovat. Rick Lovat testified that this was not undertaken because Lovat anticipated that the results of an additional analysis, as suggested by Wardrop, would be more "conservative", that is, the predicted levels of deflection would be even lower than those indicated by Wardrop's original analyses.

[30] Moreover, according to Rick Lovat, the relative deflection of the TBM components could be readily calculated and, in fact, was calculated by Lovat and Wardrop in December 1992. Rick Lovat believed that these additional calculations -- referred to by the parties as part of an "overlay exercise" concerning the Wardrop reports -- permitted Lovat to predict the extent to which differential deflection would occur between the key elements of the TBM and, hence, whether the seal gap of the TBM would widen or narrow under assumed conditions.

[31] No notes or other written record of the overlay exercise or of the resulting differential deflection calculations were kept. [page196] Consequently, these important data were not available for review by any of the committees involved in the design process or by the Hatch engineering firm. Similarly, it appears that the Wardrop reports and the Wardrop calculations were not reviewed by any of the design committees. Nor were any detailed dimension drawings of the TBM's sealing system provided to the committees or Hatch.

(3) Failure and repair of the TBM

[32] On December 28, 1993, after the TBM had been operational for about two months, a potential problem in the sealing system was detected. To investigate the suspected problem, the TBM was diverted to a location at which the cuttinghead could be removed from the tunnel for inspection by means of a vertical shaft constructed for that purpose.

[33] On inspection of the cuttinghead, extensive damage to the sealing system was discovered, particularly in the outer areas of the seal ring, due to metal-to-metal contact and other degradation processes. Curiously, however, not all the seals were crushed or otherwise compromised. Nonetheless, soil and other contaminants had infiltrated the main bearing of the TBM, and the sealing system as a whole had failed. The TBM's "Maginot Line" had been overcome. The trial judge stated: "[I]t is clear that the seals overall failed to perform as intended."

[34] Lovat repaired the TBM by introducing several design changes at a relatively low cost. These included the addition of a bronze "wear ring" to the bulkhead of the TBM to prevent the seals from being crushed in the future. Tunnelling then recommenced and eventually was completed on December 8, 1994. Subsequent re-examination of the cuttinghead revealed that excess differential deflection had occurred again, but without significant damage to the seals.

(4) Causation and foreseeability

[35] Causation and the forseeability of the failure of the TBM were critical issues at trial. Both sides called engineering experts to provide opinion evidence about the nature, cause and foreseeability of the failure of the TBM.

[36] Dr. Leslie G. Hampson testified for the CN companies. He was accepted as an expert in several engineering-related fields including, especially, methodologies for the investigation of the failure of mechanical equipment and tribology (the science of lubrication, friction and wear applicable to machine components such as bearings, seals and gears). Dr. Hampson has specific [page197] experience with TBM sealing problems. He is not an expert, however, in structural engineering or finite element analysis.

[37] Dr. Norbert K. Becker, the insurers' engineering expert, was qualified at trial as an expert in structural engineering and failure investigation. In contrast to Dr. Hampson, Dr. Becker has no particular expertise in tribology, sealing systems, tunnelling or TBMs. Dr. Becker is qualified, however, to perform finite element analyses and to interpret output data from such analyses.

[38] Both Drs. Hampson and Becker prepared a series of reports regarding the design and the cause and foreseeability of the failure of the TBM, and both testified at trial. There was considerable common ground between their causation opinions. In particular, they both agreed that the damage occasioned to the TBM was caused by excess differential deflection. The trial judge indicated in his reasons [at paras. 62 and 65]:

With respect to faulty or improper design, both experts also agree that the physical evidence of damage to the seals and other components of the TBM must lead to the conclusion that there had been excess differential deflection as between the plenum and the bulkhead. . . . . .

The failure with which we are concerned in the case at bar is the failure of the TBM to operate as intended so as to prevent soil or other foreign material from passing through the seal gap and entering into the bearing chamber thereby creating a risk of damage to the bearing. As to the cause of the failure, the totality of the evidence before this court, and particularly the evidence of the experts, Dr. Hampson and Dr. Becker, is that the damage to the seals and the wear on components of the TBM indicating metal-to-metal contact between the rotating component and the stationary component indicate that the seal gap had closed at certain points below the permitted tolerance and that such closure is consistent with excess differential deflection.

[39] Drs. Hampson and Becker disagreed, however, on whether the cause of the failure of the TBM -- excess differential deflection -- was foreseeable. In a report dated May 21, 2003, Dr. Hampson offered the following opinions:

Underlying causes of the failure

24. It was and remains my opinion that the essential engineering cause of the failure was small amounts of differential structural deflection of the cutter-head [sic] in the area of the sealing elements which proved incapable of withstanding that deflection. . . . . . . . .

25. In my opinion, differential deflection in the seal area could not have been foreseen at this time and in this situation. Identifying such a 'needle in a haystack' requires knowledge both that it is necessary and also knowing where to focus attention. With hindsight, this is always clear. . . . . . . . . [page198]

26. It is unrealistic to expect that every possible failure mode will always be anticipated and avoided especially when advancing technology -- if so there would never be any failures in service. What happened in this case is that the TBM failed by a mechanism that was not anticipated and was outside previous experience. . . . . . . . .

27. The St. Clair TBM is not a situation in which the potential failure mechanism was predicted and the external conditions were more severe than expected -- hence causing the failure. It is the opposite in that the external conditions were accurately predicted but the failure mechanism could not be foreseen and the established and successful design process did not aim to deal with it.

(Emphasis added)

[40] At trial, Dr. Hampson revised his opinion on foreseeability as expressed in para. 30 of his May 2003 report, quoted above. In an exchange with the trial judge, he acknowledged that "differential deflection is inevitable", and that it was incorrect to say that differential deflection could not have been foreseen. Dr. Hampson claimed, however, that what could not have been foreseen was excess differential deflection sufficient to "[close] the seal gap to or below the critical amount in all the seal locations". [^4]

[41] It was Dr. Becker's opinion that the TBM's sealing system failed due to differential deflection arising because the structure of the TBM's cuttinghead lacked sufficient rigidity to maintain the seal gap necessary to protect the main bearing of the TBM. In his view, this lack of rigidity was an "inherent structural design defect in the cuttinghead".

[42] Dr. Becker strongly disagreed with Dr. Hampson's opinion that the excess differential deflection was not foreseeable. Dr. Becker maintained that Lovat's engineers "knew or ought to have known" that the TBM's sealing system was vulnerable to failure because of the "enormous loads" to be transmitted through the cuttinghead, which would cause deflections "that would reduce the running gap required for these seals". Dr. Becker claimed that the failure of the TBM could have been averted if an appropriate seal ring had been installed in the TBM from the outset. He also said that the consequences of a sealing system failure could have been mitigated if the design of the TBM had accommodated in situ repairs. [page199]

[43] According to Dr. Becker, these defects established that the structural engineering of the TBM fell below the applicable structural engineering standard, with the result that the design of the TBM was flawed.

[44] Drs. Hampson and Becker both testified that a combination of circumstances, including excess differential deflection, could have led to the failure of the TBM's sealing system. There was no evidence at trial, however, that any circumstances other than excess differential deflection occurred.

IV. Trial Judge's Decision

[45] The trial judge held [at para. 66]:

[T]he cause of the failure of the TBM was excess differential deflection between components of the TBM which the TBM could not accommodate and which resulted in damage to the sealing system permitting soil and other foreign material to enter the main bearing area of the TBM.

[46] When summarizing his principal holdings at the end of his reasons, the trial judge expressed his causation conclusion this way (at para. 174): "The cause of the failure of the TBM was excess differential deflection between components of the TBM which the TBM was not capable of withstanding."

[47] The trial judge rejected the insurers' claim that the failure of the TBM was caused by faulty or improper design within the meaning of the Policy. In his view, the applicable jurisprudence in Ontario required, as a prerequisite to any holding that a design was "faulty" or "improper", the finding that the design failed to "accommodate or provide" for a condition or occurrence that was foreseeable, however unlikely or remote (at para. 68). He held that the cause of the failure of the TBM -- excess differential deflection between components of the TBM that the TBM was not capable of withstanding -- was not foreseeable, that "the design of the TBM was such as to accommodate all foreseeable risks" and that, accordingly, the design of the TBM was neither faulty nor improper within the meaning of the faulty or improper design exclusion (at paras. 76 and 174).

[48] The trial judge reached a similar conclusion regarding the inherent vice exclusion. He held that this exclusion related strictly to the peril of a vice inherent or incidental to the insured property itself, as distinct from loss occasioned by an extraneous or "adventitious" cause (at para. 58). He accepted the CN companies' submission that the insurers had not discharged their onus of establishing the inherent vice exclusion (at para. 59).

[49] The trial judge also held that, if he was incorrect in concluding that the cause of the failure of the TBM was not foreseeable, [page200] with the result that the exclusion for faulty or improper design applied, the CN companies could not recover for resultant loss or damage (at para. 99). He rejected the CN companies' submission that the faulty or improper design exclusion applied only to the cost of making good the faulty design of the TBM -- that is, to the cost of inspecting, repairing and modifying the TBM -- and not the peril of faulty or improper design itself. In the trial judge's view, properly read, the faulty or improper design exclusion applied to all loss or damage caused by the faulty or improper design of the TBM, including the inspection, repair and modification costs incurred by the CN companies (at paras. 81, 94 and 174).

[50] Finally, the trial judge rejected the CN companies' claim that the temporary blockage of the tunnel while repairs and modifications were being made to the TBM constituted physical loss or damage to the tunnel itself, within the meaning of the Policy. Rather, the positioning of the TBM in the tunnel was a deliberate act taken in the course of construction of the tunnel. Accordingly, it was not a risk of direct physical loss or damage to insured property for which coverage was provided under the Policy (at paras. 105, 106 and 174).

V. Issues

[51] As posited by the insurers during oral argument before this court, there are two issues in the appeal:

(1) Did the trial judge err by holding that excess differential deflection was not a foreseeable risk and that design error was not established at trial, with the result that the faulty or improper design exclusion was inapplicable?

(2) In the alternative, did the trial judge err in his interpretation of the inherent vice exclusion?

[52] Five issues arise in the cross-appeal:

(1) Did the trial judge err by applying a "standard of foreseeability", rather than a "standard of reasonable foreseeability", to his construction of the faulty or improper design exclusion?

(2) Did the trial judge err in his interpretation of the scope of the faulty or improper design exclusion, or in his interpretation of the resultant loss or damage exception to this exclusion?

(3) Did the trial judge err by finding that there was no physical loss or damage to the tunnel occasioned by the failure of the TBM? [page201]

(4) Did the trial judge err in his quantification of the CN companies' damages and costs?

(5) Did the trial judge err in his calculation of prejudgment interest?

VI. Analysis

A. Appeal

(1) Faulty or improper design exclusion

[53] The crux of the insurers' appeal regarding the faulty or improper design exclusion is an attack on the trial judge's foreseeability analysis.

[54] In their factum, the insurers argued that the trial judge erred by purporting to apply a "standard of foreseeability" in deciding whether the design of the TBM was faulty or improper. They maintained that this standard has no relevance to the construction of a faulty or improper design exclusion and that, in any event, the standard actually applied by the trial judge was a "standard of reasonable foreseeability". They next submitted, on various grounds, that the trial judge erred by holding that the cause of the failure of the TBM was not foreseeable and that the design of the TBM accommodated all foreseeable risks.

[55] During oral argument before this court, however, the insurers pursued only the second of these grounds of appeal. They argued that the trial judge's foreseeability findings and, hence, his decision concerning the applicability of the faulty or improper design exclusion are unsustainable because: (i) on this record, the risk of the type of failure of the TBM that materialized was both foreseeable and foreseen by Lovat; (ii) certain of the background facts relied upon by the trial judge to anchor his overall foreseeability findings are not supported by the record; (iii) the trial judge erred by failing to distinguish between the foreseeability of the mechanism by which the TBM might fail and the foreseeability of the risk of such failure; and (iv) the trial judge's approach to foreseeability was inconsistent with the nature and purpose of an "all risks" builders risk insurance policy. We will address each of these submissions in turn.

(i) Application of the standard of foreseeability

[56] As we have said, during oral argument, the insurers essentially abandoned their challenge to the trial judge's decision to apply a "standard of foreseeability" in his construction of the faulty or improper design exclusion. The appropriateness of this [page202] standard, however, is also an issue in the cross-appeal. [^5] Accordingly, while we address this issue later in these reasons, we make these preliminary observations.

[57] The trial judge correctly recognized that the insurers bore the onus at trial of establishing the application of an exclusion under the Policy. His reasons indicate that he was also aware that the courts have applied different standards in the construction of faulty design exclusions for the purpose of determining, in a given case, whether the faulty design of insured property was made out. The trial judge characterized these varying standards in these terms: (i) the "prima facie" standard, which contemplates that the fact of the failure of the property in question establishes that the design of the property was flawed; (ii) the "reasonable foreseeability" or tort standard, which mandates an inquiry into whether all reasonably foreseeable risks were taken into account in the preparation of the design; and (iii) the "foreseeability" standard, which requires that the design in question provide for all foreseeable risks.

[58] In Ontario, the Superior Court adopted a standard of foreseeability in the construction of faulty design exclusions in Foundation Co. of Canada Ltd. v. American Home Assurance Co. (1995), 1995 7102 (ON SC), 25 O.R. (3d) 36, [1995] O.J. No. 2164 (Gen. Div.), affd [1997] O.J. No. 2332, 71 A.C.W.S. (3d) 957 (C.A.) and Algonquin Power (Long Sault) Partnership v. Chubb Insurance Co. of Canada, [2003] O.J. No. 2019, [2003] O.T.C. 446 (S.C.J.). As expressed by the trial judge in Foundation at pp. 47-48 O.R., this standard requires that "All foreseeable risks must be taken into account in a design." See also Algonquin at paras. 176 and 179. In contrast, a prima facie standard has been endorsed by courts in other jurisdictions: see Queensland Government Railways and Electric Power Transmission Pty., Ltd. v. Manufacturers' Mutual Insurance, Ltd., [1969] 1 Lloyd's L.R. 214 (H.C.A.); Willowbrook Homes (1964) Ltd. v. Simcoe and Erie General Insurance Co., 1980 ABCA 139, [1980] A.J. No. 855, 22 A.R. 15 (C.A.); B.C. Rail Ltd. v. American Home Assurance Co., 1991 5713 (BC CA), [1991] B.C.J. No. 697, 79 D.L.R. (4th) 729 (C.A.).

[59] In this case, the trial judge adopted the foreseeability standard enunciated in Foundation and Algonquin as appropriate to [page203] the interpretation of the faulty or improper design exclusion, holding (at para. 54):

I . . . conclude that the law of Ontario is that the standard to be applied to determine whether a design was faulty or improper is that insured property must be designed so that it accommodates all foreseeable risks, even though such risks may be unlikely and remote, and that the standard is not the prima facie standard . . . nor a standard of reasonable foresee-ability or negligence.

[60] Elsewhere in his reasons, the trial judge reiterated this conclusion in these terms (at para. 174):

The applicable standard to determine whether there is faulty or improper design of the TBM is that the TBM must be designed to withstand all foreseeable risks in the sense of being foreseeable even if unlikely or remote. The standard is not a prima facie standard nor a reasonably foreseeable or negligent standard.

And, at para. 68:

I adopt the conclusion reached by Wilson J. in Foundation, supra, that in order to find that a design was faulty or improper, one would have to find that the design failed to accommodate or provide for a condition or occurrence that was foreseeable, however unlikely or remote.

See also paras. 55, 59, 66 and 76 of the trial judge's reasons.

[61] In the above-referenced passages from his reasons, the trial judge repeatedly stated that to establish a faulty or improper design, the foreseeability standard that he accepted requires a showing that the design in issue failed to "accommodate" all foreseeable risks. He also referred to the relevant design "provid[ing] for" or "withstand[ing]" such risks. We do not regard these descriptors as departing in any material way from the requirement expressed by the trial judges in Foundation, supra, and Algonquin, supra, that all foreseeable risks "be taken into account" in the challenged design.

[62] In our view, the trial judge's formulation of the foreseeability standard properly recognized that satisfaction of this standard requires proof that all foreseeable risks have been identified and addressed in the design in question. Mere recognition of a foreseeable risk is insufficient. "Accounting" for a foreseeable risk contemplates both that the risk is identified and that provision or allowance is made in the impugned design to meet the identified risk. On the foreseeability standard, anything less will not establish a fault-free and proper design. Nor, in our opinion, does designing against a foreseeable risk convert the risk into an unforeseeable one. It simply means that the applicable design provided for the risk, that is, the risk was identified and addressed in the design with a view to forestalling its occurrence, [page204] thus meeting the foreseeability standard. In this context, we agree that the foreseeability standard mandates that the relevant design "take into account", "accommodate", "provide for" and "withstand" all foreseeable risks.

[63] Having accepted this foreseeability standard as appropriate to the interpretation of the faulty or improper design exclusion, the trial judge then applied this standard to the facts of this case, holding that the cause of the failure of the TBM was not foreseeable and that the design of the TBM "was such as to accommodate" all foreseeable risks (at para. 76). These key foreseeability findings precluded reliance by the insurers on the faulty or improper design exclusion.

[64] None of the parties to these proceedings attacked the trial judge's formulation of the foreseeability standard. Rather, as relevant to their interests, they challenged the appropriateness and manner of its application by the trial judge in this case.

[65] In their factum, for example, the insurers argued that the trial judge's adoption of this standard was misconceived. They asserted that the use of the word "faulty" in the exclusion does not reference a legal standard derived from negligence law. Rather, the language of the exclusion simply establishes the requirement that the design in question be "defective", that is, not fit for the purpose intended. Thus, in effect, the insurers urged adoption of what the trial judge termed a prima facie standard.

[66] In oral argument before this court, however, the insurers took the position that even on the standard of foreseeability endorsed by the trial judge, they had met their onus to invoke the faulty or improper design exclusion.

[67] To satisfy this onus, it is insufficient to demonstrate merely that the challenged design failed to perform as expected. It suffices to establish that the impugned design failed to provide for all foreseeable risks, including those that were reasonably foreseeable.

[68] The insurers' case on appeal in respect of the faulty or improper design exclusion rests on the claim that even the demanding standard accepted by the trial judge was satisfied, with the result that the exclusion was triggered. In these circumstances, for the purpose of the appeal, the determination of whether a less stringent standard might apply in evaluating the design of the TBM is not required: see this court's decision in Foundation, supra, at para. 3. Nor is it necessary to address the insurers' additional argument that the standard in fact applied by the trial judge was the tort standard of reasonable foreseeability. We will return to the issue of the standard of foreseeability adopted by the trial judge in our discussion of the issues in the cross-appeal. [page205]

(ii) Foreseeability of risk and design error

[69] The trial judge's penultimate holding concerning the faulty or improper design exclusion was set out at para. 76 of his reasons:

Accordingly, applying the applicable standard as developed in the recent judicial decisions in this province, it is my view that excess differential deflection to the extent necessary to cause the failure of the TBM which occurred was not foreseeable in the sense of being foreseeable even though unlikely or remote. I conclude therefore that the design of the TBM was such as to accommodate all foreseeable risks and that accordingly the failure of the TBM was not caused by faulty or improper design as referred to in the exclusion of "the cost of making good faulty or improper design" in clause I 3(d) (iii) of the Policy.

[70] As is apparent from these statements, the trial judge's conclusion that the faulty or improper design exclusion could not be relied upon by the insurers to deny coverage rests on his foreseeability findings. These findings, in turn, are predicated on and flow from the trial judge's causation analysis.

(a) Trial judge's causation analysis

[71] The overarching causation issue at trial was the determination of the cause of the failure of the TBM. Both experts agreed, and the trial judge accepted, that this failure was occasioned by the contamination of the TBM's main bearing due to the collapse of the TBM's protective sealing system.

[72] A critical component of this issue, however, concerned the cause of the failure of the sealing system itself. The trial judge recognized the significance of this issue, commenting, "The failure with which we are concerned in this action is the failure of the sealing system of the TBM . . . The major issue to be determined is the cause of such failure" (at para. 24) (emphasis added).

[73] After reviewing the evidence of both Drs. Hampson and Becker on the issue, the trial judge concluded "on all of the evidence before [the] court", that the cause of the failure of the TBM was excess differential deflection that the TBM could not accommodate (at para. 66) or withstand (at para. 174).

[74] In reaching this conclusion, the trial judge reasoned in part [at paras. 55, 56, 62 and 65]:

In the case at bar, there is no evidence of any possible cause of the failure of the TBM other than excess differential deflection. The question becomes whether the elimination of all other possible causes of the failure establishes that the failure must be due to excess differential deflection. . . .

The only evidence before the court as to possible causes of the failure is the evidence of the two expert witnesses who were consistent in that the only possible causes of failure could be faulty or improper design, inherent vice, [page206] faulty or inadequate maintenance, faulty operation, overpressurization of the plenum or overheating of the seals. Both experts agreed that there was no evidence of the latter four possible causes. . . . . .

With respect to faulty or improper design, both experts also agree that the physical evidence of damage to the seals and other components of the TBM must lead to the conclusion that there had been excess differential deflection as between the plenum and the bulkhead. . . . . . . . .

As to the cause of the failure, the totality of the evidence before this court, and particularly the evidence of the experts, Dr. Hampson and Dr. Becker, is that the damage to the seals and the wear on components of the TBM indicating metal-to-metal contact between the rotating component and the stationary component indicate that the seal gap had closed at certain points below the permitted tolerance and that such closure is consistent with excess differential deflection.

(Emphasis added)

[75] The trial judge's reasons demonstrate that he was also alive to the evidence of Drs. Hampson and Becker that some combination of circumstances could have led to the compromise of the TBM's sealing system. At para. 66 of his reasons, the trial judge expressly adverted to Dr. Hampson's evidence on this issue, but then commented:

I am not aware of any evidence before the court indicating that any circumstance other than excess differential deflection had occurred although both experts agree that it is difficult to explain how excess differential deflection caused damage to certain of the seals.

[76] The trial judge elaborated at para. 67:

Both [experts] concede that there could be some combination of circumstances, which combined with the excess differential deflection, resulted in the failure. Even if there was some combination of circumstances including the excess differential deflection, in the absence of any evidence of any unpredicted or unusual external cause or condition, some combination of circumstances must all have been circumstances relating to the TBM itself. Accordingly, one must conclude, on the balance of probabilities, that the design of the TBM was deficient in that it was not able to perform as specified under expected or normal conditions.

[77] Thus, as the insurers maintain, the trial judge found that there was no evidence of a cause of the failure of the TBM independent of the TBM itself. To the contrary, the evidence pointed exclusively to excess differential deflection as the root cause of the failure.

[78] Accordingly, the trial judge's holding that the TBM failed due to excess differential deflection that it could not "accommodate" and that it was not "capable of withstanding" when the [page207] TBM's sealing system was overcome, was amply grounded in the expert evidence at trial. To this extent, there is no basis for appellate interference with the trial judge's holdings.

[79] That, however, does not end the matter. The difference of opinion between the engineering experts was a narrow but central one: it concerned whether excess differential deflection was foreseeable when the TBM was designed. The trial judge indicated at para. 67:

The critical difference of opinion between Dr. Hampson and Dr. Becker is that it is Dr. Hampson's view that "the differential deflection in the seal area could not have been foreseen at this time and in this situation" whereas, it is Dr. Becker's view that the lack of rigidity causing the excess differential deflection was the result of a design defect that should have been foreseen by Lovat. . . . The difference of opinion as between the experts is whether the failure was foreseeable.

[80] As we read his reasons, the trial judge saw little or no difference between the opinions of the engineering experts as to the cause of the failure of the TBM. We agree. As a result, the real issue in respect of the faulty or improper design exclusion concerns the trial judge's pivotal findings that "excess differential deflection to the extent necessary to cause the failure of the TBM which occurred was not foreseeable" and that "the design of the TBM was such as to accommodate all foreseeable risks".

(b) Trial judge's foreseeability findings

[81] The CN companies argue that the trial judge's foreseeability findings are unassailable. We disagree. For four reasons, we accept the insurers' contention that the trial judge's foreseeability findings are tainted by reversible error.

Failure to consider Rick Lovat's evidence on foreseeability of risk

[82] First, in undertaking his foreseeability analysis, the trial judge considered in detail the evidence of Drs. Hampson and Becker regarding the foreseeability of the cause of the failure of the TBM. He accepted Dr. Hampson's opinion that excess differential deflection compromising of the TBM's sealing system was not foreseeable. In doing so, however, the trial judge failed to consider Rick Lovat's evidence on the foreseeability of this failure risk.

[83] In the course of summarizing Dr. Hampson's evidence, the trial judge alluded to the finite element analyses performed by Wardrop to determine the anticipated deflection of specific components of the TBM -- including the cuttinghead and the bearing [page208] assembly -- under predicted pressures, and the results of those analyses (at para. 26). He also referred to Dr. Hampson's evidence about the Wardrop analyses and the "overlay exercise" conducted by Lovat and Wardrop in relation to the Wardrop results (at paras. 23 and 27), and to Dr. Becker's testimony regarding the sufficiency of the Wardrop modelling (at paras. 28 to 30). Later in his reasons, the trial judge outlined the insurers' position that the Wardrop analyses were deficient or inaccurate and that they signified that Lovat was aware of the risk of excess differential deflection (at para. 74) and the response of the CN companies to these claims (at para. 73). He did not review, however, Rick Lovat's evidence on these issues.

[84] It is undisputed before this court that Lovat was aware when it designed the TBM that: (i) the main bearing of the TBM was a critical component of the TBM; (ii) it was essential to maintain the TBM's seal gap within the specific design tolerance of six millimetres if the main bearing was to be insulated from contaminants; (iii) soil and other contaminants presented a serious threat of damage to the main bearing if they came into contact with that bearing; and (iv) the structural integrity of the TBM was solely Lovat's responsibility.

[85] It is against this unchallenged factual backdrop that Rick Lovat's testimony concerning the risk of excess differential deflection and the steps taken by Lovat to address that risk must be considered.

[86] Rick Lovat was Lovat's lead representative on the tunnelling project. He was cross-examined extensively at trial on what Lovat knew at the time that the TBM was designed, and on the purpose, nature and results of the Wardrop analyses and the overlay exercise. He confirmed that, in its finite element analyses, Wardrop investigated whether any of the modelled TBM components would deflect under anticipated pressures and, if so, to what degree. During cross-examination, he said:

Q. And again what they were looking for was whether or not any of the components they modelled would deflect, correct?

A. Correct.

Q. And, in particular, whether they would actually deform plastically, bend, or dent to the point where they wouldn't recoil?

A. And if they deflected, by how much.

(Emphasis added)

[87] Later in his cross-examination, Rick Lovat acknowledged that if the seal flange of the TBM deflected at a different rate than the fixed bulkhead, differential deflection affecting the critical [page209] seal gap would occur. He also conceded that if the deflection of the seal flange exceeded the deflection of the fixed bulkhead, the seal gap would narrow.

[88] In respect of the overlay exercise, Rick Lovat testified that it caused him to believe that Lovat was able to predict "the extent to which differential deflection between the [modelled] elements was occurring" (emphasis added). Importantly, he also said that, by virtue of the overlay exercise, Lovat specifically addressed its mind to the prospect that differential deflection could occur that would damage the TBM's sealing system by narrowing the seal gap. On cross- examination, this telling exchange took place:

Q. Okay. Now, and just so that we are clear on this, Mr. Lovat, do I understand, am I correct that in doing this analysis you have just told me about, these additional calculations, Lovat was in fact attempting to determine whether there would be differential deflection and, if so, what its effect on the seal gap would be?

A. Correct.

Q. Okay. So Lovat actually turned its mind to the issue of whether or not the seal gap would be maintained within tolerances under anticipated loads and expected operating conditions?

A. Correct.

Q. All right.

A. I wouldn't necessarily say that there would be expected operating conditions, but under the conditions that we had provided Wardrop. . . . . .

A. We had generated a case which was far more conservative than the actual operating case.

Q. In fact more force than what you thought you would encounter?

A. Correct.

Q. All right. So worse case scenario?

A. Yes.

Q. All right. So having turned your mind to the very issue of will differential deflection occur and adversely affect the seal gap, I take it then, sir, based on the calculations you say Lovat and Wardrop did, Lovat satisfied itself that differential deflection, which might occur, would not be a sufficient magnitude to make the sealing gap move outside of design tolerances?

A. Correct.

(Emphasis added)

[89] These admissions indicate that Lovat knew when the TBM was designed that differential deflection posed a serious potential danger to the TBM's sealing system. More importantly, [page210] the extent of this potential danger was investigated by Lovat and Wardrop. As a result, it was Lovat's opinion that the magnitude of such deflection would not compromise the sealing system. Although this opinion may have been understandable on the basis of Wardrop's modelling and Lovat's interpretation of the overlay exercise calculations, this does not detract from the fact that, according to Rick Lovat, this type of risk was both known to and investigated by Lovat or its agents. Indeed, during his testimony, Dr. Hampson confirmed that he had been informed by Rick Lovat that as a result of the Wardrop analyses and the overlay exercise, Rick Lovat had concluded that the seal gap of the TBM would be maintained at a "satisfactory level".

[90] At one point in his reasons, the trial judge noted that the Wardrop analyses were conducted to "determine the extent of differential deflection" (at para. 75). Significantly, however, he made no mention of Rick Lovat's testimony, quoted above, in his foreseeability analysis. Nor does the record otherwise support the conclusion that the trial judge assessed the significance of this evidence in making his foreseeability findings. Instead, the trial judge's foreseeability analysis was focused on the expert engineering evidence about foreseeability and excluded any consideration of the import of Rick Lovat's evidence on this fundamental issue.

[91] In our view, with respect, this was an error. In Keljanovic Estate v. Sanseverino, 2000 5711 (ON CA), [2000] O.J. No. 1364, 186 D.L.R. (4th) 481 (C.A.), at pp. 489-90 D.L.R., leave to appeal to S.C.C. refused [2000] S.C.C.A. No. 300, a majority of this court characterized this type of error as a "processing error" [at para. 30]:

The second kind of error that may warrant appellate interference is what might be called a "processing error", that is an error in processing the evidence that leads to a finding of fact. This type of error arises when a trial judge fails to appreciate the evidence relevant to a factual issue, either by disregarding or misapprehending that evidence. When the appellate court finds such an error it must first determine the effect of that error on the trial judge's reasoning. It may interfere with the trial judge's finding if it concludes that the part of the trial judge's reasoning process that was tainted by the error was essential to the challenged finding of fact.

See also Waxman v. Waxman, 2004 39040 (ON CA), [2004] O.J. No. 1765, 186 O.A.C. 201 (C.A.), at para. 334, leave to appeal to the S.C.C. refused [2004] S.C.C.A. No. 291.

[92] We recognize that the trial judge was not required to refer to all the evidence in his reasons and that his factual findings, absent palpable and overriding error, attract deference from this court. The trial judge's lengthy reasons demonstrate an impressive command of the trial record, and a detailed consideration of [page211] much of the evidence. Nor should we be taken as suggesting that the trial judge failed to consider any of Rick Lovat's evidence, or the ample evidence establishing that the possibility of differential deflection was an issue of which the TBM design participants were aware.

[93] But differential deflection is not the same as differential deflection of the type or magnitude sufficient to overtake the sealing system of the TBM, that is, excess differential deflection compromising of the TBM's sealing system. The trial judge expressly found that the latter risk was not foreseeable. We do not read his reasons as indicating that he was satisfied that the risk of such excess differential deflection was identified and taken into account in the design of the TBM. To the contrary, we read his reasons as holding precisely the opposite: this form of risk was not foreseeable or foreseen, with the result that the faulty or improper design exclusion was not triggered.

[94] Yet, Rick Lovat's evidence bore directly on this issue and, hence, on material findings of fact that led the trial judge to conclude that the faulty or improper design exclusion did not apply. His evidence was highly relevant to the question of whether the failure of the TBM's sealing system due to excess differential deflection was foreseeable and foreseen and, thus, to the ultimate question of whether the faulty or improper design exclusion applied.

[95] In these circumstances, Rick Lovat's testimony could not be disregarded. We conclude that the trial judge's failure to consider the import of this evidence taints his foreseeability findings with palpable and overriding error warranting appellate intervention.

Misapprehension of evidence

[96] Second, the trial judge's foreseeability findings appear to have been based, at least in part, on a misapprehension of the evidence concerning the technical committees' and Lovat's knowledge of the risk of failure of the TBM due to excess differential deflection. In this respect, as well, the trial judge erred in processing the evidence leading to his critical foreseeability findings. We say this for the following reasons.

[97] In holding that the cause of the failure of the TBM -- excess differential deflection that overcame the TBM's sealing system -- was not foreseeable, the trial judge relied on his earlier factual finding that the Wardrop analyses and Lovat's review of those analyses "appear to have been accepted by the [design] committees as addressing any concerns with respect to differential deflection" (at para. 75). [page212]

[98] Contrary to this finding, however, there was no evidence at trial that any of the technical committees or Hatch reviewed the results of the overlay exercise. The record indicates that only those Lovat and Wardrop personnel who took part in the overlay exercise saw these differential deflection calculations. Moreover, there was no evidence at trial demonstrating that the committees evaluated the Wardrop analyses, or the risk of failure of the TBM's sealing system due to excess differential deflection. For example, the minutes of the meeting of the technical committee contain no reference to any discussion about the possible deflection of the cuttinghead of the TBM. As well, the trial judge found that the design committees asked very few questions about the TBM's sealing system or its potential for differential deflection (at para. 69). In these circumstances, the trial judge's conclusion that the design committees were satisfied by the Wardrop analyses and the overlay exercise with respect to differential deflection concerns is not supported by the record.

[99] In addition, at para. 75 of his reasons, the trial judge relied on Dr. Hampson's suggestion, set out in his May 2003 report, that:

[C]uttinghead differential deflection and the potential effect on the sealing elements had not been a previous problem and was not identified at the time by anyone as a potential problem. The assembled expertise had no reason to anticipate this new failure process.

(Emphasis added)

[100] The evidence, however, established that the Wardrop analyses and the overlay exercise were intended to address the potential for damaging differential deflection on various components of the TBM, including the cuttinghead and the seal gap. Thus, Dr. Hampson's assertion that the potential effect of differential deflection on the TBM's sealing elements "was not identified at the time by anyone as a potential problem" was inaccurate and the trial judge's reliance on it was misplaced. Contrary to Dr. Hampson's claim, this potential risk was both recognized and investigated by Lovat -- the designer of the TBM.

[101] The unsupported finding of the trial judge, set out above, and his endorsement of the quoted passage from Dr. Hampson's report appear in the trial judge's reasons immediately before his articulation of his dispositive findings on foreseeability. In our view, it is clear that they formed part of the underpinnings for those findings. This part of the trial judge's reasoning process was both tainted by error and essential to his foreseeability findings. As observed by this court in Waxman, supra, at para. 335, a [page213] finding of fact in the absence of any evidence is a serious processing error, which constitutes a palpable error.

Conflation of foreseeability risks

[102] Third, we also agree with the insurers that the trial judge erred by failing to distinguish between the foreseeability of the type of risk of failure that occurred here and the foreseeability of the mechanism by which such risk might materialize. This error went to the heart of the trial judge's forseeability analysis.

[103] As we have already indicated, relying on Foundation, supra, and Algonquin, supra, the trial judge concluded that the determination of whether the design of the TBM was faulty or improper required an inquiry into whether the TBM was designed so that it accommodated or provided for all foreseeable risks, even though such risks might be unlikely or remote. At para. 67 of his reasons, he correctly stated that, under this standard, the relevant question was whether "the risk which caused the failure" was foreseeable. Ultimately, the trial judge held at para. 76 that "excess differential deflection to the extent necessary to cause the failure of the TBM which occurred was not foreseeable in the sense of being foreseeable even though unlikely or remote".

[104] The trial judge appears to have based this holding on Dr. Hampson's revised evidence at trial, given in response to a question by the trial judge, that while differential deflection was both inevitable and foreseeable, the scale of the differential deflection that occurred was not foreseeable. During Dr. Hampson's testimony, the following interchange took place:

The Court: And it would be correct to say it's not -- it's not right that differential deflection could not have been foreseen, what could not have been foreseen was differential deflection sufficient to, I think as you put it, collapse all the seals?

The Witness: Close all the seal gaps, or take -- let me be specific because I haven't been specific in the way I should -- of closing the seal gap to or below the critical amount in all the seal locations.

[105] In this exchange, Dr. Hampson conceded, apparently for the first time, that the type of risk of damage to the TBM's sealing system that occurred in this case -- damage from differential deflection -- was foreseeable. However, he also shifted the focus of his foreseeability opinion away from the foreseeability of the type of risk involved, to the foreseeability of the scale of that risk. By accepting this nuanced shift in Dr. Hampson's evidence, the trial judge conflated the foreseeability of the type of risk that could cause the failure of the TBM with the [page214] foreseeability of the magnitude of this risk. In so doing, in our opinion, the trial judge erred.

[106] The issue of the scale of the risk of differential deflection was an inquiry into whether the mechanism by which differential deflection might damage the sealing system was foreseeable, rather than an inquiry into whether it was foreseeable that deflection compromising to the seal gap might occur at all. However, the controlling issue to be determined at this trial was whether the type of risk of failure of the TBM that materialized was foreseeable. If it was not, the insurers could not rely on the faulty or improper design exclusion to deny coverage.

[107] In fairness to the trial judge, given Dr. Hampson's evidence, this confusion may well have arisen because Dr. Hampson repeatedly emphasized the foreseeability of the "potential failure mechanism" of the TBM. [^6] By accepting Dr. Hampson's evidence that the TBM failed by a mechanism that was not anticipated, the trial judge failed to focus on the foreseeability of the failure risk in issue -- the risk of differential deflection affecting the integrity of the TBM's sealing system. The failure "mechanism" or event ultimately described by Dr. Hampson in his evidence was not the risk of such damaging differential deflection but, rather, the combination of that risk with the inadequacy of the measures taken to address or accommodate it in the design of the TBM. While that combination may have been unforeseeable, this is not tantamount to a conclusion that the risk itself was unforeseeable.

[108] The CN companies rely on Foundation, supra, to argue that the type of risk of failure at issue in this case was not foreseeable. The facts in Foundation, however, are not analogous to the facts of this case.

[109] In Foundation, during the construction of a substructure on a bridge, a cofferdam on the bridge experienced a blow-in (the sudden erosion or collapse of part of the river rockbed in the vicinity of the cofferdam), which was caused by a pocket of gas and slickenside. The applicable insurance policy contained an exclusion concerning "the cost of making good faulty or defective . . . design". The trial judge found as a fact that there was no experience or knowledge in the engineering community when the dam was designed regarding the combined effect of slickenside [page215] and gas. Rather, the combined presence of gas and slickenside was an unpredictable, unknown and rare event. As a result, this combination of circumstances was not taken into account in the design of the dam. [^7]

[110] In contrast, in this case, the technical committees and Lovat knew that structural steel can deflect under various loads and pressures. Lovat and Wardrop recognized the potential danger of differential deflection to the cuttinghead and the seal gap. They undertook analyses the admitted purposes of which included the determination of whether differential deflection that could exceed the design tolerance of the TBM's seal gap, thereby narrowing the gap and endangering the main bearing, would occur under worse case conditions. Thus, the potential occurrence of the causative event of the failure of the TBM was anticipated.

[111] Moreover, unlike the facts in Foundation, the trial judge in this case explicitly found that there was no evidence of an external cause of the failure of the TBM. To the contrary, he held that even under the unproven thesis that some combination of causes accounted for the failure of the TBM, on the evidence before him these necessarily related, not to some external causative event, but to the TBM itself.

[112] Finally, the design of the TBM was modified after the failure to protect the sealing system against the effects of excess differential deflection. In contrast, in Foundation, when the dam was rebuilt after the blow-in, no alterations to its design were made.

[113] It is also useful to contrast Foundation with Algonquin, upon which the insurers rely. In Algonquin, the trial judge held that an exclusion for "faulty, inadequate or defective design" applied to losses occasioned when a hydroelectric dam was damaged by hydraulic piping, that is, by the development of pipes or channels leading to soil erosion and the formation of sinkholes. She found that the risk of such damage was well known in the industry, that the dam "did not fulfill the purpose for which it was intended; it was not designed to take into account the foreseeable risk of piping" (at para. 182), and that faulty design was the cause of the failure of the dam (at para. 200). Thus, the type [page216] of risk that materialized in Algonquin, unlike the type of risk that occurred in Foundation, was not unexpected, rare or unknown.

[114] Like Algonquin, the type of risk that materialized in this case -- the risk that the TBM's protective seal gap could be compromised by differential deflection leading to corruption of the critical main bearing -- was known. Furthermore, by virtue of the analyses that it had conducted or commissioned, Lovat believed that this known risk was adequately addressed by the design of the TBM. Thus, the type of risk at issue here was both foreseeable and foreseen. As it happened, Lovat was mistaken in its opinion that the TBM's sealing system could withstand the differential deflection that might occur. Consequently, this risk was not "accommodated" by the design of the TBM, in the sense that the design lacked those features or attributes necessary to meet this foreseeable risk. Instead, the design of the TBM was not fit for its intended purpose and the failure of the TBM was the result of a foreseeable and foreseen causative event. The design of the TBM, therefore, was faulty within the meaning of the Policy.

[115] Importantly, the trial judge found as a fact that "[T] he design of the TBM was deficient in that it was not able to perform as specified under expected or normal conditions" (at para. 67) and that, "Although all of the seals were not damaged, it is clear that the seals overall failed to perform as intended" (at para. 15). It is also telling, as noted by the trial judge, that Dr. Hampson himself acknowledged that the failure of the TBM was the result of "inherent features of the TBM not fulfilling their intended function" (at para. 27).

[116] The standard of foreseeability endorsed by the trial judge required that all known and foreseeable risks of failure be addressed in the design of the TBM. This did not occur. Instead, contrary to the findings of the trial judge, the design of the TBM was inadequate to meet a known risk of possible failure of the TBM. We are compelled to conclude, therefore, that the design of the TBM was faulty and, consequently, that the faulty or improper design exclusion applies.

Nature of insurance policy at issue

[117] We complete our consideration of the trial judge's treatment of the faulty or improper design exclusion with this observation. An "all risks" policy of property insurance, like any other insurance policy, represents an agreed allocation between an insurer and an insured of those risks that the insurer is prepared to underwrite and those that are to be borne by the insured. [page217] Insurance policies of this type provide broad coverage for losses and damage to property. But they do not provide coverage against all conceivable perils. Obviously, were it otherwise, there would be no role for coverage exclusions.

[118] The Policy here was neither a warranty that the insured property -- the TBM -- would fulfill its intended purpose, nor a warranty of entrepreneurial design risk. With respect, the effect of the trial judge's foreseeability analysis in this case is to convert the Policy into just such a warranty.

[119] For these reasons, we conclude that the faulty or improper design exclusion was made out.

(2) Inherent vice exclusion

[120] In support of their denial of coverage under the Policy, the insurers also relied on the exclusion in the Policy for inherent vice. For convenience, we again set out the terms of this exclusion:

3.EXCLUSIONS: This Policy does not insure: . . . . .

(e) Wear, tear, inherent vice, normal upkeep and normal making good; but this exclusion shall not apply to resulting loss not otherwise excluded by this Policy.

(Emphasis added)

[121] The trial judge found that the insurers had not made out the inherent vice exclusion because there was no evidence to support a finding of inherent vice (para. 61). He appeared to accept the definition of inherent vice from Employers Casualty Co. v. Holm, 393 S.W. 2d 363 (Tex. Civ. App. 1965), at p. 367:

The term "inherent vice" as a cause of loss not covered by the policy, does not relate to an extraneous cause but to a loss entirely from internal decomposition or some quality which brings about its own injury or destruction. The vice must be inherent in the property for which recovery is sought.

[122] We would not disturb the trial judge's conclusion that the insurers did not make out the inherent vice exclusion. That exclusion has been interpreted narrowly by the courts, most recently by the British Columbia Court of Appeal in Dawson Creek (City) v. Zurich Insurance Co., 2000 BCCA 158, [2000] B.C.J. No. 463, 75 B.C.L.R. (3d) 131 (C.A.), leave to appeal to S.C.C. refused [2000] S.C.C.A. No. 217. In that case, the insured sought compensation for damages caused by collapse of the roof of an arena. The evidence showed that the collapse was due to the combination of weakness of the structure from faulty construction and the snow load on the day of the collapse. The snow load was not out of the ordinary. Relying on [page218] C.C.R. Fishing Ltd. v. British Reserve Insurance Co., 1990 145 (SCC), [1990] 1 S.C.R. 814, [1990] S.C.J. No. 34, a marine insurance case, the court held that the loss fell within the policy because it would not have occurred but for the faulty construction which was "fortuitous in that it was not to be expected in the ordinary course of things" (Dawson Creek at para. 36).

[123] In C.C.R., the marine policy excluded loss from "inherent vice or nature of the subject matter insured". In that case, a ship sank either because someone had left a valve open or because repairers had used the wrong type of cap screws, which had corroded. Justice McLachlin held that even if the loss was due solely to the corroded cap screws, this was not inherent vice. She said the following at p. 821 S.C.R.:

The concept of inherent vice in the context of marine insurance refers to loss stemming from qualities inherent in the thing lost. The failure of the cap screws in the case at bar cannot be said to result from purely inherent qualities of the ship. The unfortunate installation of these parts in the ship was a result of the negligence of the repairers, an external cause unrelated to those qualities. The loss was fortuitous, in the sense that it was not the inevitable product of a quality inherent in the vessel.

(Emphasis added)

[124] Other cases have given a similarly narrow interpretation to inherent vice. For example, Soya G.m.b.H Mainz Kommanditgesellschaft v. White, [1983] 1 Lloyd's L.R. 122 (H.L.), at p. 126: "It means the risk of deterioration of the goods shipped as a result of their natural behaviour in the ordinary course of the contemplated voyage without the intervention of any fortuitous external accident or casualty."

[125] Thus, inherent vice refers to some internal decomposition that made deterioration of the thing inevitable because of qualities inherent in the TBM. The insurers adduced no evidence to that effect, and simply relied on the inference to be drawn from the fact that the TBM failed. They argued, in effect, that since the TBM failed under normal conditions there must have been some inherent defect, if the failure was not the result of faulty or improper design. The burden was on the insurers to prove the inherent vice exclusion. The trial judge was not prepared to draw the conclusion, simply from the failure of the TBM, that the failure was due to inherent vice. Given the narrow interpretation of the inherent vice exclusion, we cannot find that the trial judge erred in this approach, given his finding that the cause was excess differential deflection.

[126] We would not give effect to this ground of appeal. [page219]

B. Cross-Appeal

(1) Application of the standard of foreseeability

[127] As the insurers originally did in these proceedings, the CN companies challenge the trial judge's application of a standard of foreseeability to the construction of the faulty or improper design exclusion. Unlike the insurers, however, the CN companies maintain that the relevant standard is the tort standard of reasonable foreseeability. They argue that this court's decision in Foundation, supra, permits this court to apply this standard and that, on this record, its application supports the conclusion that all reasonably foreseeable risks of the TBM's failure were adequately addressed in the design of the TBM. In our opinion, this argument cannot succeed for two reasons.

[128] First, we do not accept that a faulty or improper design exclusion in an "all risks" builders risk insurance policy is to be construed in accordance with a negligence-based reasonable foreseeability standard. We agree with the views expressed by the trial judge in Foundation, supra, at p. 47 O.R., quoted with approval in Algonquin, supra, at paras. 175-77 and 179 that the finding of a design error or a defective design may, but need not, depend upon a showing of designer negligence. In this case, on its face, the faulty or improper design exclusion is concerned with shortcomings in the attributes or character of the insured property itself, as designed, rather than the conduct of the designer.

[129] Perhaps more importantly, the operative question is whether the impugned design accounted for all foreseeable risks, rather than only those that were reasonably foreseeable. As observed by Lang J. in Algonquin at para. 176:

[I]f the design of the structure was fit for its intended purpose, and the designer had accounted for all potential foreseeable risks, then the insurer would be unable to satisfy a court that the design was faulty within the meaning of the policy's exclusion. However, if the designer had only accounted for reasonably foreseeable risks, then this would not be sufficient, and the insurer could prove that the faulty design exclusion applied.

(Emphasis in original)

[130] On this approach, as indicated in Algonquin at para. 179, an insurer will succeed in establishing faulty design if it demonstrates that "the participants failed to design [the property in question] so that it would function for its intended purpose after accounting for all foreseeable risks".

[131] Second, and in any event, the facts of this case do not support the CN companies' contention that all reasonably foreseeable [page220] risks of the TBM's failure were adequately addressed in the design of the TBM.

[132] The CN companies acknowledge that the failure of the TBM was caused by the collapse of its sealing system. We do not understand them to dispute the trial judge's finding that this occurred due to excess differential deflection. We have already concluded that the risk of excess differential deflection compromising of the TBM's sealing system was known to and investigated by Lovat or its agents when the TBM was designed. Thus, in contrast to the facts in Foundation but like the facts in Algonquin, the causative risk of failure in this case was not only reasonably foreseeable, it was a known risk that was actually foreseen by the designer itself. In the result, however, the steps taken to meet that risk were insufficient.

[133] In this context, we reiterate that the trial judge found that the TBM "was not able to perform as specified under expected or normal conditions" and, consequently, "the design of the TBM was deficient" (at para. 67). Moreover, the TBM's sealing system "failed to perform as intended" (at para. 15). Thus, this case involves a design of insured property that was insufficient to permit the insured property to function for its intended purpose, after the investigation of a known and predictable potential peril but the failure to introduce any design changes to meet that peril. In these circumstances, we have no hesitation in concluding that the faulty or improper design exclusion is triggered.

(2) Resulting loss or damage exception

[134] In their cross-appeal, the CN companies submit that the trial judge erred by holding that if the exclusion for faulty or improper design applied, they could not recover for resultant loss or damage. This submission turns on the exception to the faulty or improper design exclusion set out in clause 3(d)(iii) of the Policy:

3.EXCLUSIONS: This Policy does not insure: . . . . .

(d) the cost of making good . . . . .

(iii) faulty or improper design

provided, however, to the extent otherwise insured and not otherwise excluded under this Policy, resultant loss or damage under any Section of this Policy shall be insured . . . ;

(Emphasis added) [page221]

[135] The CN companies argue that if any resultant loss or damage is insured under any other parts of the Policy, they are entitled to recover for that loss or damage. For example, they point to Section II of the Policy, which insures against loss resulting from delay beyond the scheduled start-up date caused by "a peril insured against under Section I of this Policy". They claim that the interaction between Sections I and II permits them to claim damages caused by the delayed start-up that resulted from the faulty or improper design of the TBM. The trial judge rejected this argument. He held as follows at para. 99:

In my view, this provision must be construed to mean that coverage for all claims made under the Policy for loss or damage resulting from faulty or improper design, if not claims for physical loss or damage under Section I of the Policy, must be found under specific provisions of other Sections of the Policy and must not be excluded under specific provisions of other Sections of the Policy.

[136] The CN companies point out that the resultant loss or damage clause is worded differently from clauses of this type considered by the courts of this country. One of the leading cases, and a case relied upon by the trial judge, is British Columbia v. Royal Insurance Co. of Canada, 1991 5732 (BC CA), [1991] B.C.J. No. 2959, 60 B.C.L.R. (2d) 109 (C.A.), leave to appeal to S.C.C. refused [1991] S.C.C.A. No. 506. In that case, the relevant clause read as follows [at p. 111 B.C.L.R.]:

This policy does not insure

(a)(i) faulty or improper material;

(ii) faulty or improper workmanship;

(iii) faulty or improper design;

provided, however, to the extent otherwise insured and not otherwise excluded under this policy, resultant damage to the property shall be insured . . .

(Emphasis added)

[137] Thus, unlike this case, which uses the phrase "resultant loss or damage under any Section of this Policy", in Royal Insurance the policy referred to "resultant damage to the property". The CN companies submit that by using the words "under any Section of this Policy" and omitting the words "to the property", the parties eliminated the requirement that there must be damage to other property. Rather, resultant damage under any section of the Policy, such as delayed opening, is covered. The CN companies argue that the trial judge, in effect, erred by inserting the words "to the property" after the word "damage" in clause 3(d)(iii) of the Policy. [page222]

[138] We agree with the trial judge that the resulting loss or damage exception to the faulty or improper design exclusion does not provide coverage. The CN companies' submission ignores the basic structure of the Policy. Section I of the Policy insures against all risks of "direct physical loss or damage" subject to the exclusions in clause 3. One of those exclusions is "the cost of making good . . . faulty or improper design". We have found that the loss or damage to the TBM was the result of faulty or improper design, which comes within the exclusion in clause 3(d)(iii). This physical loss or damage is not otherwise insured under any other part of the Policy. Therefore, by reason of the exclusion of the peril for faulty or improper design, there is no "peril insured against under Section I" within the meaning of Section II of the Policy. As a result, there is no coverage, for example, for the loss or damage caused by the delayed start-up, within the resulting loss or damage exception to the faulty or improper design exclusion. We agree with the insurers that the delayed opening insurance under Section II provides coverage for consequential economic loss only to the extent that the economic loss results from insured loss or damage to insured property.

[139] We do not agree that this interpretation renders the exception in clause 3(d)(iii) meaningless. The insurers provide an example. If the faulty design of the TBM had caused a fire that damaged the TBM and part of the tunnel constructed behind it, the damage to the TBM would be excluded, as would any economic loss from delayed opening due to the damage to the TBM. However, the physical damage to the tunnel itself, which was insured property, would be covered, as would any economic loss from delayed opening attributable to the damage to the tunnel.

[140] In a slight variation on this submission, the CN companies maintain that the faulty or improper design exclusion applies only to the cost of making good the faulty design and not to the peril of faulty design. To put this submission in concrete terms, the CN companies assert that the only cost excluded is the cost incurred by Lovat for the inspection, repair and modification of the TBM cuttinghead. The trial judge rejected this submission. As he said at para. 81, "[T]he exclusion must be intended to apply to all loss or damage caused by faulty or improper design and is not limited to the costs incurred in repairing or modifying the faulty or improper design of the TBM." The trial judge analyzed the two American decisions upon which the CN companies relied and upon which they rely in this court. We agree entirely with the trial judge's complete analysis at paras. 78-94 of his reasons for judgment. [page223]

(3) Sue and labour and salvage

[141] The CN companies also rely upon the sue and labour provision in Section IV of the Policy. As we indicated earlier, clause 16 of Section IV provides that in case of loss or damage "it shall be lawful and necessary for the Insured . . . to sue, labour and travel for in and about the defence, safeguard and recovery of the property insured hereunder" and "the Insurer(s) will contribute . . . towards any reasonable and proper expenses in connection with the foregoing". The CN companies submit that this clause provides coverage for the expense of digging the vertical shaft and extracting the TBM, being costs necessarily incurred to prevent any further damage to the property. For example, they argue that if the TBM had continued to tunnel and failed under the St. Clair River, this would have caused irreparable damage to the tunnel.

[142] The CN companies make a similar submission in relation to Statutory Condition 9, which requires the insured to "take all reasonable steps to prevent further damage to any such property so damaged and to prevent damage to other property insured hereunder". Statutory Condition 9 requires the insurer to contribute toward any reasonable and proper expenses in connection with the steps taken by the insured.

[143] In our view, the decision of the Supreme Court of Canada in Hartford Fire Insurance Co. v. Benson & Hedges (Canada) Ltd., 1978 37 (SCC), [1978] 2 S.C.R. 1088, [1978] S.C.J. No. 53 determines these issues against the CN companies. Speaking for the majority at p. 1102 S.C.R., Pratte J. held that these clauses cover mitigation of damage resulting "from the contingency that has occurred as opposed to being the consequence of another contingency that has yet to occur". Further, as Lang J. said in Algonquin Power, supra, at para. 154:

Such clauses oblige an insurer to contribute to reasonable and proper expenses incurred to prevent further damage. They do not, when worded as in the Chubb/Partnership policy, apply to indemnify the insured for expenses incurred to minimize or mitigate an uninsured loss.

[144] Thus, the submissions based on these clauses fail because the contingency of failure under the St. Clair River never materialized and because they do not apply to cover expenses incurred to minimize an excluded loss.

(4) Physical loss or damage to the tunnel

[145] In their cross-appeal, the CN companies also argue that the tunnel was physically damaged and that this constitutes direct physical loss under Section I of the Policy. This submission [page224] is based on the theory that while the tunnel was blocked by the TBM it could not function as a tunnel. The trial judge fully dealt with this argument at paras. 100-08 of his reasons and we agree with his analysis. As he said at para. 105: "[T]he positioning of the TBM in the tunnel was not a fortuitous event but was clearly a deliberate act taken in the course of construction of the tunnel. Such deliberate act does not . . . constitute a risk of direct physical loss or damage to insured property for which coverage is provided under the Policy."

[146] Further, as the insurers point out, the interpretation of this part of Section I is assisted by reference to clause 3(a), which provides that the Policy does not insure "any loss of use or occupancy howsoever caused except as specifically stated". Thus, the Policy itself distinguishes between direct physical loss or damage and loss of use. Loss of use of the tunnel is addressed in Section II, which deals with expenses for delayed opening. But, as we have already said, by reason of the exclusion of coverage for faulty or improper design, the coverage in Section II is not triggered.

[147] Therefore, this ground of cross-appeal also fails.

(5) Other grounds of cross-appeal

[148] As we have indicated, the CN companies also challenge the quantum of damages and costs awarded by the trial judge, and his calculation of prejudgment interest. These issues, however, are predicated on the entitlement of the CN companies to recover under the Policy. Because we have concluded that the faulty or improper design exclusion operates to preclude such recovery, it is unnecessary to address these remaining grounds of cross-appeal.

VII. Disposition

[149] Accordingly, for the reasons given, the appeal is allowed and the cross-appeal is dismissed. The trial judgment dated October 7, 2004 is set aside and the action against the appellant insurers is dismissed. The insurers shall submit their brief written submissions concerning the costs of these proceedings, and of the trial if sought, to the Registrar of this court within 20 days from the date of this decision. The CN companies shall submit their brief responding written submissions on costs to the Registrar within 20 days thereafter. [page225]

LANG J.A. (dissenting): --

Overview

[150] I have had the opportunity to read the thorough and careful reasons of my colleagues. In particular, I have read the reasons explaining their conclusion that the trial judge erred in fact and in law in his decision interpreting and applying the faulty design exclusion contained in this "all- risks" insurance policy. Respectfully, I am unable to agree with their analysis.

[151] In my view, this case turns primarily on the interpretation and application of the faulty design exclusion and the onus on the insurer to prove that exclusion. Subject to one clarification, in my view, the trial judge interpreted and applied the test correctly. Since the insurers failed to satisfy the onus on them to establish the exclusion, I would dismiss the appeal.

[152] This was a complex case that consumed 19 days of trial time. The trial judge reserved judgment for more than five months and wrote 176 paragraphs of reasons supporting his conclusion. It is apparent that the trial judge had an excellent understanding of the technical facts and the considerable scientific evidence presented at trial. In my view, the trial judge's factual findings are supported by the evidence and are entitled to deference. To the extent that my colleagues find that the trial judge misapprehended certain evidence, I cannot agree. However, even if I accepted the errors found by my colleagues, in my view, those errors would not change the result.

[153] Regarding the appropriate test, in my view, the governing standard is that adopted by this court in Foundation Co. of Canada Ltd. v. American Home Assurance Co. (1995), 1995 7102 (ON SC), 25 O.R. (3d) 36, [1995] O.J. No. 2164 (Gen. Div.), affd [1997] O.J. No. 2332, 71 A.C.W.S. (3d) 957 (C.A.). That standard requires the designer to take all foreseeable risks into account. As was stated in Foundation, the resolution of this issue necessitates a comparative analysis of the design under attack against some standard.

[154] However, in my view, my colleagues' reasons alter the Foundation standard in two important ways. First, the standard as formulated in their reasons requires not only that all foreseeable risks be taken into account in the design, but also that the design succeed in accommodating those risks. Second, their reasons do not consider the essential comparative component of the Foundation test. In making these two alterations to the standard, my colleagues adopt a standard that is significantly closer to the results-focused standard set out in [page226] Queensland Government Railways and Electric Power Transmission Pty., Ltd. v. Manufacturers' Mutual Insurance Ltd., [1969] 1 Lloyd's L.R. 214 (H.C.A.). The Queensland standard was rejected by this court in Foundation.

[155] Regarding the application of the standard to the facts of this case, as I read his reasons, particularly at para. 76, the trial judge found that the insured machine was designed to take into account all foreseeable risks. The insured assembled a world-wide committee of experts to determine and approve the conceptual design for the machine and a second committee of experts to oversee its manufacture. To this end, the insured then contracted with a very experienced manufacturer, Lovat Tunnel Equipment Inc. (Lovat), that had successfully manufactured 124 other tunnel boring machines, none of which experienced the failure that occurred in this case. The manufacturer, who was responsible for the structural integrity of the machine, undertook analyses to ensure that the design took into account all foreseeable risks. The trial judge did not accept any evidence called by the insurer that indicated that any further steps should have been taken in the design process to forestall the risk that materialized. Accordingly, as I understand the evidence and the trial judge's reasons, all foreseeable risks were taken into account.

[156] I turn first to a consideration of the law, which in my view, is the crux of this appeal. On the law, the CN companies ("CN") argued that the trial judge properly applied the foreseeability test adopted by this court in Foundation. Although the insurers' oral argument, in contrast to their factum, focused on the application of the test, from my perspective, the test must be correct before it can be applied to the relevant facts.

The Insurance Policy

[157] To begin, the interpretation of the faulty design exclusion must be considered in accordance with the relevant principles of interpretation of an insurance contract.

Principles of insurance interpretation

[158] Importantly, the trial judge was mindful of and thoroughly considered the general principles applicable to the interpretation of an insurance contract. He recognized that, once the insured proves the loss, the onus shifts to the insurer "to establish on a balance of probabilities that the loss comes within an exclusion in the policy". The trial judge also noted that "exclusion provisions are to be construed narrowly against the insurer". The trial judge appreciated that, when faced with ambiguity, such as [page227] the meaning of faulty design, a court should seek an interpretation consistent with commercial reality and the true intent of the parties in accordance with the purpose of the coverage and of the exclusion.

[159] He recognized that coverage is only provided for fortuitous losses in the sense of being random or unexpected and that, accordingly, there is no coverage for losses flowing from ordinary wear and tear or from the inevitable degradation of the object insured. The same principle negates coverage for losses wilfully caused by the insured. This is because insurance protects against risks, not certainties. However, a policy may provide coverage for losses arising from the insured's negligence, such as liability insurance. See Consolidated Bathurst Export Ltd. v. Mutual Boiler and Machinery Insurance Co., 1979 10 (SCC), [1980] 1 S.C.R. 888, [1979] S.C.J. No. 133; Craig Brown, Introduction to Canadian Insurance Law, 2nd ed. (Markham, Ont.: LexisNexis Butterworths, 2006).

[160] Cognizant of these broad insurance principles, the trial judge turned to the interpretation of this particular policy, including its purpose.

The purpose of the policy

[161] Two pre-conditions were essential to CN's tunnel project. First, CN needed a tunnel boring machine that would dig a very large tunnel. Second, CN needed insurance to protect it in case that machine failed, the tunnel collapsed, or some other calamity befell the project. In negotiating the $890,000 premium for this large-scale tunnelling project, the insurers and the insured recognized that the two primary risks were the potential failure of the TBM and of the tunnel or its lining. Because the risks were high, and the potential losses significant, the risk was spread among six insurers.

[162] CN was worried about and investigated the risks of this project. Relevant to the risk that manifested in this case, Klohn Leonoff, a specialized engineering firm retained by CN, commented on the TBM proposal at the time the contract was signed in 1992. Those comments referred to "Unexpected Events", including "Failure of Main Bearing". The relevant comment said that such a failure is "feared" because the consequences would be expensive. The comments refer to the need for elasticity in the seals given the larger diameter of this TBM and refer to solving the problem by arranging the two groups of seals in different directions, as implemented by Lovat. In addition, the insurers were aware that the main bearing was the piece of equipment most likely to fail and that the main bearing was "the critical [page228] component because it is exposed to the bored material first and thus most susceptible to damage".

[163] With these concerns in mind, CN and the insurers negotiated the terms of coverage through the services of a broker.

The terms of insurance

[164] The policy agreed upon, which CN bought in March 1993, not only insured against specific perils such as fire and flood, but against all risks, including soil conditions and the mechanical breakdown of the TBM. As such, it provided extremely broad coverage, subject only to the specified exclusions, such as the faulty design exclusion and the custom exclusions required by the insurers.

[165] Specifically, the policy provided coverage for "ALL RISKS of direct physical loss or damage . . . to . . . all . . . property . . . including but not limited to the tunnel boring machine", as well as the structures, tunnels and materials used during construction. The TBM was the key and most expensive element of this project and was recognized as such by the insurers. As I have said, the risk of main bearing failure from exposure to bored material, which I take to include soil, was explicitly considered by the insurers.

[166] In addition to the faulty design exclusion, the insurance contract contained carefully considered exclusions commensurate with what the insurer saw as the unacceptable risks. For example, the insurers specifically excluded coverage for mechanical breakdown. However, this exclusion was limited to the mechanical breakdown of "any belt, screen, grinding and/ or cutting blade, disc or surface, or any wear plate or mantle which forms part of any machine".

[167] The insurers did not exclude the mechanical breakdown of the main bearing.

[168] The policy also excluded certain risks specific to the tunnel boring machine, including depletion, deterioration, corrosion, wear and tear and the leakage of particular named parts of the machine -- none of which were at issue in this case. Importantly, the insurers did not exclude the risk of leakage of soil through the seals, whether caused by excess differential deflection ("EDD") [^8] or otherwise. [page229]

[169] Because the policy did not contain an exclusion for the TBM's main bearing breakdown, nor for leakage through the seals, nor for even the common exclusion of losses resulting from "any external cause", such as unexpected soil conditions, it would seem the insurers were prepared to bear these risks, subject to the exclusion for faulty design.

[170] With this context, I turn to the question of the interpretation of the faulty design exclusion.

The Faulty Design Exclusion [^9]

The test

[171] In my view, it is important to revisit the origins and development of the faulty design exclusion foreseeability test because it is the formulation of this test that is determinative of this appeal.

[172] Historically, to establish the applicability of a faulty design exclusion, the insurer was required to show negligence on the part of the designer. This changed beginning with Queensland, a decision of the High Court of Australia. In that case, piers collapsed because they were designed with a prismatic shape, a shape that proved inadequate to withstand an unanticipated level of flood waters. The court held that the insurer proved faulty design by establishing only that the design of the piers was defective for its intended purpose. While designer negligence might assist in establishing that the design was defective, the exclusion applied even if the designer met all professional standards. Thus, the Queensland standard is results-focused -- the structure failed therefore its design must be faulty.

[173] The Alberta Court of Appeal in Willowbrook Homes (1964) Ltd. v. Simcoe and Erie General Insurance Co., 1980 ABCA 139, [1980] A.J. No. 855, 22 A.R. 95 (C.A.) adopted the Queensland standard and excluded coverage, even though construction site walls were braced in accordance with industry standards. Coverage was excluded because the bracing did not work for its intended purpose in the face of inordinate wind pressure.

[174] Collavino Inc. v. Employers Mutual Liability Co. of Wisconsin, 1984 5963 (ON SC), [1984] O.J. No. 1011, 7 C.L.R. 165 (H.C.J.), affd [1985] O.J. No. 227, 14 C.C.L.I. xli (C.A.), similarly excluded coverage on the Queensland standard for a trestle that failed to withstand [page230] normal spring ice flows; however, the insurers also established that the designers ignored the recommendation of their consulting engineers to investigate for ice pressure.

[175] In B.C. Rail Ltd. v. American Home Assurance Co., 1991 5713 (BC CA), [1991] B.C.J. No. 697, 79 D.L.R. (4th) 729 (C.A.), a railway track embankment collapsed because it could not be supported by the underlying soil. The British Columbia Court of Appeal adopted Queensland and denied coverage because the embankment failed to work for its intended purpose; however, the insurer also established that the designer made incorrect assumptions regarding the composition of the soil because it failed to perform any soil tests.

[176] In Foundation, the trial judge rejected both the Queensland standard and the standard of reasonable foreseeability, concluding instead that the correct standard is one of foreseeability. In coming to this conclusion, Wilson J. noted at p. 46 O.R. that cases such as Willowbrook, Collavino and Queensland all shared a common element: "the design of the structure itself was inadequate to meet the demands of extreme, but foreseeable, circumstances". In those circumstances, there was evidence to support the finding that the design was defective because the foreseeable event was not taken into account.

[177] The facts in Foundation involved the collapse of a cofferdam caused by the collapse of the underlying river bed, which in turn was caused by a "blow-in" resulting from a rare combination of slickenside and gas. In these circumstances, Wilson J. explained her reasons for rejecting a standard of reasonable foreseeability at p. 47 O.R.:

The departure from the test of negligence as a prerequisite for faulty or more particularly defective design makes sense and I accept it. It would be too easy for design standards to slip, and be dictated by economic factors. Short cuts could be taken at both the insurer's expense, and at the expense of principles of public safety. Meeting the standard of the profession at the time is not sufficient to avoid the exclusion.

[178] In the result, Wilson J. set the appropriate standard as one of simple foreseeability, a standard that requires the insurer to establish that the designer failed to take into account all foreseeable risks.

[179] In applying this foreseeability standard, Wilson J. noted that the cause of the failure, the rare combination of slickenside and mud, was "unforeseen", in that it was unlikely further tests, such as another bore hole, "would have identified the potential problem", and thus all foreseeable risks had been taken into account in the cofferdam's design. She found no fault with the design, but only that the design had been unable to withstand [page231] the unforeseeable risk. She reasoned that in these circumstances, to exclude coverage when all foreseeable conditions were taken into account, would render the "all-risks" insurance "illusory".

[180] In reaching her conclusion, Wilson J. determined that the foreseeability standard necessarily requires a comparison of the design to some standard and, since on the evidence before her, it was "most unlikely" that further tests "would have identified the potential problem", the insurer failed to meet its onus to prove the exclusion.

[181] In addressing the appropriate comparator, Wilson J. said at p. 47 O.R.:

I pause to note that a finding of faulty or defective design is necessarily comparative. It means comparing the design in question to some standard, and requires a finding that the design was below that standard. The issue of whether the standard in the industry is the appropriate one is at the heart of the dilemma of defining the appropriate test. Public safety is affected by the standards that are set. . . . . .

The test cannot be simply the prevailing industry standard . . . A design that is in accordance with industry standards is of course persuasive to rebut an allegation of faulty or defective design. It is not however determinative of the issue. I recognize that in many cases there will be an overlap between the standard imposed for faulty design and the standard of negligence. However the two tests are not necessarily identical. All foreseeable risks, rather than reasonably foreseeable risks, must be taken into account in the design. Reasonable foreseeability leaves too great a potential margin for error, as judged by Queensland and B.C. Rail.

(Emphasis added)

[182] On appeal, this court endorsed Foundation's foreseeability standard, determined that the Queensland standard did not apply in Ontario, and left for another day the question of whether the more forgiving reasonable foreseeability standard should apply.

[183] Algonquin Power (Long Sault) Partnership v. Chubb Insurance Co. of Canada, [2003] O.J. No. 2019, [2003] O.T.C. 446 (S.C.J.), was the next Ontario case to consider the foreseeability standard. In that case, which involved the collapse of a hydroelectric dam, the insurer succeeded in excluding coverage because it established that adverse soil conditions of hydraulic piping in the dam's embankment were predictable. Since the designer did not take into account this foreseeable risk by undertaking appropriate testing, the design was faulty and coverage was excluded. In adopting the Foundation test, Algonquin looked at whether the designer had taken into account or considered all foreseeable risks and confirmed the need for a comparative analysis. [page232]

[184] After canvassing the standard as set out in Foundation and Algonquin, the trial judge in the instant case concluded at para. 54:

I adopt the reasons of Wilson J. in Foundation, supra, and Lang J. in Algonquin, supra, and conclude that the law of Ontario is that the standard to be applied to determine whether a design was faulty or improper is that insured property must be designed so that it accommodates all foreseeable risks, even though such risks may be unlikely and remote, [^10] and that the standard is not the prima facie standard as enunciated in Queensland, supra, nor a standard of reasonable foreseeability or negligence.

Interpretation of standard

[185] As I read my colleagues' reasons, they do not compare the conduct of the designer to any standard and they conclude that the faulty design exclusion relates to the "shortcomings in the attributes or character of the property itself, as designed, rather than the conduct of the designer". The critical part of my colleagues' reasoning is found at para. 114 of their reasons, which states:

As it happened, Lovat was mistaken in its opinion that the TBM's sealing system could withstand the differential deflection that might occur. Consequently, this risk was not 'accommodated' by the design of the TBM, in the sense that the design lacked those features or attributes necessary to meet this foreseeable risk.

[186] In the same vein, my colleagues refer to the "inadequacy of the measures taken to address or accommodate [excess differential deflection] in the design of the TBM".

[187] For me, these extracts, and my colleagues' reasons generally, set a standard akin to Queensland, which I cannot accept as the governing standard.

[188] I deal first with the adoption of "accommodate" as an added element to the Foundation standard. This addition originated with the trial judge, who used the term in his reasons when he stated, "I adopt the reasons of Wilson J. in Foundation, supra, and Lang J. in Algonquin, supra, and conclude that the law of Ontario is that the standard to be applied to determine whether a design was faulty or improper is that insured property must be designed so that it accommodates all foreseeable risks" (emphasis added). As pointed out by my colleagues, in other [page233] paragraphs of his reasons, the trial judge uses "provide" and "designed to withstand", both to the same effect. The problem with this language is that neither Foundation nor Algonquin used these terms in delineating the standard.

[189] In my view, the trial judge simply misspoke when he used these words rather than "take into account". I say this for four reasons.

[190] First, I conclude that "accommodate" was unintentional because the trial judge substituted that word or "provide" in the foreseeability standard without explanation. In my view, had he intended to change the standard so significantly, he would have explained his reasons for doing so.

[191] Second, in stating that the insured property must be designed to accommodate all foreseeable risks, the trial judge indicated that he was adopting the standard set out in Foundation and Algonquin, neither of which employs the word "accommodate".

[192] Third, in the same sentence that incorporates "accommodate", the trial judge specifically confirms his rejection of the Queensland standard. A standard of failure to "accommodate" would be very close to the results-focused Queensland standard, which has been rejected by this court.

[193] Fourth, the addition of the word "accommodate" is inconsistent with the trial judge's application of the standard to the facts of this case.

[194] Accordingly, in my view, the trial judge did not intend to change the standard. Such a high standard that looks only at outcome, without any consideration of the conduct of the designer, would ignore principles of interpretation that require a narrow construction of exclusions and would result in an interpretation that significantly raises the bar in favour of the insurer. I say this because, if a design is required to succeed in accommodating all foreseeable risks, as my colleagues say it must, then the design is required to meet a standard of perfection with respect to those risks.

[195] Rather, in my view, the governing standard only requires the insured to take all foreseeable risks into account. This is consistent with the Foundation standard that calls for a comparative analysis, an analysis that considers the human component of the design, and asks what the designer anticipated, or should have anticipated, would be the risks arising from the design and whether the designer took those risks into account. If the insurer can show, on a balance of probabilities, that the designer failed to meet this standard, the insurer will have succeeded in proving the exclusion. If the insurer fails to meet that onus, the insured is entitled to coverage. [page234]

[196] The foreseeability standard is more favourable to the insurer than the reasonable foreseeability standard because it sets a higher standard to which the design must conform. At the same time, the foreseeability standard achieves the goal of providing insurance coverage commensurate with the designer and the insured's reasonable expectations. It also ensures that an insurer is not required to compensate an insured for any shoddy design or cost-cutting measures employed by the designer, even if those measures met industry standards.

[197] Further, the foreseeability standard strikes a fair balance in respect of allocation of risk. While such a test imposes a high standard on the insured, it is not as high as the standard of perfection implicit in "accommodation". Under my colleagues' standard, designers and their customers could be discouraged from undertaking innovative projects if faced with exclusion from coverage on a results-focused basis, irrespective of the diligence exercised in achieving the design.

[198] In this case, it is unnecessary to delineate the precise comparative standard because there was no evidence accepted by the trial judge that any other designer would have undertaken any different analysis and the facts support a finding that Lovat was a most prudent designer. There is a broad spectrum of conduct between that of a negligent designer and one who meets the most exacting of standards. The precise location of the standard on that spectrum is best determined on the appropriate factual foundation.

Application of the standard

[199] On my reading of the trial judge's reasons, he properly applied the correct standard and found that the designer of the TBM took into account all foreseeable risks. On the facts of this case, the designer recognized the risk of EDD and its potential to start a domino effect leading to main bearing failure. For this reason, it undertook all precautions to exclude the risk in the design and manufacture of the TBM. The structural integrity of the cuttinghead design was confirmed by the Wardrop and Lovat analyses. The Wardrop analyses provided detailed, highly sophisticated, computerized calculations of the anticipated deflection of the relevant components. The Lovat overlay exercise used those calculations to determine the degree to which those components would deflect towards each other (differential deflection). On the basis of these analyses, Lovat was satisfied that differential deflection would not exceed permitted tolerances so as to put at risk the effectiveness of the sealing [page235] system. In that way, Lovat accounted for the foreseeable risk of EDD in its design.

[200] The trial judge's finding that the Wardrop and Lovat analyses took into account the risk of deflection and differential deflection respectively was supported by the evidence. The Wardrop analyses were available at trial, including the detailed data provided for those analyses by Lovat. The insurers were unable to point to any error in those analyses. While the results of the subsequent Lovat overlay analysis were not recorded, there was no evidence that the analysis was conducted other than thoroughly and correctly. Importantly, the insurers, on whom the onus lay, called no evidence from an expert with credentials similar to those of Dr. Hampson, CN's expert, to show that the key Wardrop analysis was based on faulty data or that an overlay analysis conducted differently would have revealed that the design was at risk of EDD.

[201] Indeed, the trial judge accepted that the analyses undertaken established that the cuttinghead and seal design, including the allowance for differential deflection, would meet the purpose of protecting the seals. Further, Wardrop indicated that it anticipated that the results of any additional finite element analysis "with the addition of the cuttinghead and thrust bearing" would reflect increased structural integrity of the bearing plate area.

[202] The design, including allowances for differential deflection, was tried and true. The TBM had been designed to withstand greater pressures than those present in this project and, as found by the trial judge at para. 75, "the sealing system used had been proven at pressures greater than those encountered in the St. Clair River Tunneling Project". Dr. Hampson's expert evidence supported the trial judge's conclusion that the Foundation standard was met. There was no evidence accepted by the trial judge that any other designer would have undertaken any more or different tests. In addition, a similar design worked in 124 earlier tunnel boring machines manufactured by Lovat.

[203] The trial judge preferred Dr. Hampson's evidence to that of Dr. Becker, the insurers' expert, on the issue of foreseeability. As the trial judge noted, Dr. Hampson's qualifications were significantly more relevant to the issues than were Dr. Becker's. Dr. Becker was an expert only in structural engineering and failure investigation, with no expertise in tribology, seals, tunnelling or tunnel boring machines. Dr. Hampson not only had a particular expertise in tribology -- the science of lubrication, friction and wear as it applies to components such as bearings, seals and gears -- but also specific experience with sealing problems in [page236] tunnel boring machines. In addition, as found by the trial judge, Dr. Hampson was an expert "in the fields of methodology for failure investigation for mechanical equipment; performance of machines in hostile environments; extending life and performance of industrial plant; design audit . . . and machine condition monitoring". In my view, it was open to the trial judge to accept Dr. Hampson's evidence on the critical issue of whether the design took all foreseeable risks into account and to reject Dr. Becker's opinion concerning the advisability of an annular ring or the need to provide for in situ repairs.

[204] My colleagues refer to the post-breakdown modifications. In my view, these modifications do not serve to distinguish Foundation and cannot serve as evidence that the risk was not taken into account, particularly in light of the expert evidence to the contrary. On this point, the trial judge accepted Dr. Hampson's evidence that the modifications could not have been anticipated as important to the original design and that there was no indication that the original design "was constrained by cost or that corners were cut" in any way. Accordingly, the success of remedial measures undertaken post-breakdown cannot inform whether the breakdown was foreseeable at the relevant time.

[205] My colleagues also distinguish Foundation on the basis that the cofferdam's failure in that case occurred as a result of external causes, while the cause of soil contamination in this case was internal to the design. In my view, there should not be a different result for unforeseeable internal causes than for unforeseeable external causes. As in Foundation, it was unlikely in this case that further tests would have identified the potential problem. Since, in both this case and in Foundation, the insured took into account all foreseeable risks, the standard was met.

[206] Furthermore, the trial judge's interpretation of this exclusion is consistent with the other terms of this custom- made policy because the insurers were aware of the risks surrounding the TBM, could have excluded the risk that materialized and chose not to do so. In particular, the insurers chose not to list the bearing or the seals under either the mechanical breakdown exclusion or under the excluded components of the TBM.

[207] Finally, the trial judge's interpretation and application of the standard in this case is consistent with the onus on the insurer to prove the exclusion and with the principle of interpreting exclusion clauses narrowly and in keeping with the purpose for which the insurance was acquired. To interpret "faulty design", an ambiguous term, as meaning any design defect, whether that defect was sufficiently taken into account [page237] or not, would run contrary to the reasonable expectations of the parties.

[208] In summary, it was open to the trial judge to accept Dr. Hampson's evidence and to conclude that the design of the TBM contemplated and accounted for differential deflection and EDD. [^11] There was no evidence accepted by the trial judge that supported a finding that any other designer or manufacturer would have undertaken further or different tests for EDD.

[209] For the above reasons, and because I agree with my colleagues that the insurers failed to establish the inherent vice exclusion, I would dismiss the appeal.

[210] On a final point, I do not agree with my colleagues that the trial judge misapprehended the evidence or made any "processing" errors. However, I have not found it necessary to address these issues because, in my view, even on the facts relied upon by my colleagues, the insurers failed to meet the onus on them to establish the faulty design exclusion.

The Cross-Appeal

[211] Since I would dismiss the appeal, it is necessary for me to address only the grounds of the cross-appeal arising from such a dismissal, which include the quantum of damages relating to the Lovat setoff, costs awarded by the trial judge and the date for commencement of prejudgment interest.

The Lovat setoff

[212] In a separate action against Lovat for damages, CN alleged that the machine's breakdown was caused by Lovat's failure to design the cuttinghead with sufficient rigidity to avoid EDD. The parties attempted to settle that action. An essential term was that Lovat would buy the TBM back from CN. Immediately before the parties reached their final settlement, Lovat increased its offer by $440,000. That offer was accepted by CN. [page238]

[213] In view of this settlement, the trial judge deducted $440,000 from the amount the insurers were required to pay CN under the policy because he accepted the insurers' argument that CN would otherwise achieve double recovery; CN was not entitled to both its full claim against the insurers and the $440,000 paid by Lovat that could reasonably be attributed to the buy-back of the TBM.

[214] On its cross-appeal, CN challenges this finding on three bases.

[215] First, CN argues that there was no evidentiary foundation for the finding that Lovat's payment was attributable to the settlement of CN's action against Lovat.

[216] However, in my view, in the absence of any evidence to the contrary and in light of the surrounding circumstances, the trial judge was entitled to draw the inference that the increase in the Lovat settlement reflected an amount that both Lovat and CN attributed as compensation to CN for its claim. Indeed, the trial judge found that the $440,000 was the minimum amount that could be attributed for this purpose.

[217] Second, I am unable to agree with CN's submission that the insurers are not entitled to subrogate against this recovery until CN has fully recovered all its losses, as opposed to a full recovery of its entitlement under the insurance policy. In other words, CN argues that even if the insurers fully satisfy its insurance claim, CN is not fully indemnified because it still has losses, for example, by reason of the policy's 30-day waiting period for the delayed opening, by reason of the policy deductible, and by reason of the costs CN incurred in effecting its recovery from the insurers.

[218] As the trial judge pointed out, CN specifically contracted for a 30-day waiting period with respect to delayed opening expenses, as well as for the policy's deductible. Either one of these expenses exceeds the $440,000 at issue. In effect, CN self-insured for these losses and it is not entitled to apply against them the $440,000 it received from Lovat.

[219] Finally, the insurers do not dispute CN's argument that it is not fully indemnified under the policy until it has recovered legal costs incurred in pursuing recovery against the third party, Lovat. However, since CN led no evidence about those costs, an appropriate credit cannot be calculated.

[220] CN provided no authority to support its secondary argument that it should also be able to offset its costs incurred in pursuing recovery from the insurers against the moneys received from Lovat. In any event, I am unable to accept this argument because to do so would undermine any incentive by an insured to control litigation costs in pursuing its insurer. [page239]

Costs

[221] CN challenges the trial judge's award of costs on two bases: first, that the trial judge erred in decreasing CN's claimed costs by more than 40 per cent based on the reasonable expectations principle, and, second, that the trial judge erred in denying CN a second counsel fee for a case that was admittedly complex.

[222] In my view, the trial judge made no error on either point.

[223] First, Boucher v. Public Accountants Council for the Province of Ontario (2004), 2004 14579 (ON CA), 71 O.R. (3d) 291, [2004] O.J. No. 2634 (C.A.), clearly stands for the principle that costs awards must be subject to the "overriding principle of reasonableness". In this circumstance, even taking into account the heavier burdens placed on a plaintiff to prepare its case, the trial judge did not err in taking into consideration the fact that CN's costs for trial preparation far exceeded those of the insurers. It cannot be said that the trial judge exercised a wrong principle in applying his significant experience to the circumstances of the case and in reducing the costs to $1,150,837.35 ($800,000 of which was for fees).

[224] Second, the trial judge made no error in refusing to award a second counsel fee. The authorities are clear, while the total counsel fee may be split among several counsel, it may not exceed the maximum amount permitted by the applicable costs grid. See Celanese Canada Inc. v. Canadian National Railway Co., 2005 8663 (ON CA), [2005] O.J. No. 1122, 196 O.A.C. 60 (C.A.), at paras. 50 and 51; Walker v. Ritchie, 2005 13776 (ON CA), [2005] O.J. No. 1600, 197 O.A.C. 81 (C.A.), at para. 102, vard on other grounds [2006] 2 S.C.R. 428, [2006] S.C.J. No. 45, 2006 SCC 45; and Schaible Electric Ltd. v. Melloul-Blamey Construction Inc., 2005 26706 (ON CA), [2005] O.J. No. 3226, 201 O.A.C. 71 (C.A.), at para. 39, which confirmed that "an award for two counsel fees at trial is permissible, provided that the aggregate of the two fees does not exceed the maximum counsel fee permitted under the costs grid".

[225] Thus, I would dismiss the cross-appeal regarding costs.

Prejudgment interest

[226] CN challenges the trial judge's decision to average the prejudgment interest rate payable on this claim. In this case, the difference between the average rate and the rate set out in the Courts of Justice Act, R.S.O. 1990, c. C.43, did not fluctuate by more than 3 per cent between the date of the claim and the date of judgment.

[227] In my view, Graham v. Rourke (1990), 1990 7005 (ON CA), 75 O.R. (2d) 622, [1990] O.J. No. 2314, 74 D.L.R. (4th) 1 (C.A.) does not set down [page240] an inflexible rule precluding an averaging of rates in all circumstances where the fluctuation is less than 4 per cent and for that reason cannot be said to be "significant". Indeed, the fluctuation of the interest rate is only one factor under s. 130(1)(b) of the Courts of Justice Act, which provides a court with the discretion to "vary" the relevant rate.

[228] In this case, there were other factors to consider, particularly the lengthy delay between the denial of coverage in August 1995 and the commencement of the trial in February 2004. In these circumstances, the averaging of prejudgment interest was within the discretion of the trial judge. I see no error to justify interfering with his exercise of that discretion.

[229] As a secondary issue on this ground, CN argues that the trial judge erred in deciding that interest would commence on the date coverage was refused (August 23, 1995) rather than on the date the proof of loss for coverage was filed (March 1, 1995). The insurers did not respond to this ground of appeal.

[230] The policy specifically provides that "adjusted claims . . . shall be due and payable no more than thirty (30) days after presentation and acceptance of proofs of interest and loss at the office of the Insurer(s)". On its face, the adjusted claims were payable by the insurers to CN by March 31. Accordingly, I would allow the cross-appeal in this respect and set March 31, 1995 as the date for commencement of prejudgment interest.

Result

[231] In the result, I would dismiss the main appeal and allow the cross-appeal only to the extent of the commencement date for prejudgment interest. Since the bulk of the work related to the main appeal, I would award costs to the successful respondent.

Appeal allowed; cross-appeal dismissed.

[^1]: Many of the components of the TBM were made of structural steel, which deflects under various stress and load conditions. "Differential deflection" occurs when components of this type deflect at different rates. "Excess differential deflection" occurs when the differential deflection exceeds the permitted tolerances between the components in question.

[^2]: In 1995, St. Clair Tunnel Construction Company, a wholly owned subsidiary of Canadian National Railway Company, merged with the respondent St. Clair Tunnel Company. The merged corporation continued under the latter name.

[^3]: The term "earth pressure balance" means that this TBM was capable of partially or fully balancing the natural earth pressure to minimize settlement or heaving of the grounds above the machine.

[^4]: Later in his testimony, Dr. Hampson suggested that, in his earlier exchange with the trial judge, he had indicated that "deflection is foreseeable and inevitable". He explained that deflection can occur without any differential deflection. The transcript, however, reveals that he told the trial judge that differential deflection was both inevitable and foreseeable.

[^5]: In their cross-appeal, the CN companies maintain that the trial judge erred by failing to employ a reasonable foreseeability standard in his interpretation of the faulty or improper design exclusion. They submit that this exclusion contemplates that where the designer of insured property has undertaken everything that could reasonably be expected in the preparation of the design of the insured property, the design cannot be said to be "faulty" or "improper" and the exclusion does not apply.

[^6]: For example, in his May 2003 report, Dr. Hampson asserted that "the TBM failed by a mechanism that was not anticipated" (para. 32), that "the potential failure mechanism was [not] predicted" (para. 34), and that "the failure mechanism could not be foreseen" (para. 34).

[^7]: Applying a standard of foreseeability to the construction of the exclusion, the trial judge in Foundation held that the design of the cofferdam was fit for its intended purpose and that the blow-in was not foreseeable. Consequently, the design of the dam was not "faulty" and the exclusion did not apply. An appeal from that decision was dismissed by this court without discussion of the content of the standard of foreseeability applied by the trial judge.

[^8]: Differential deflection describes the deflection of two adjacent structures as they move towards or away from each other. The term excess differential deflection ("EDD") was used in this proceeding to describe the deflection of the two structures beyond acceptable tolerances. In this case, the two structures were the moving cuttinghead and the static bulkhead. The cuttinghead cut the soil ahead of the TBM and the bulkhead housed the main bearing.

[^9]: I do not consider any nuances raised by the additional word "improper" in this policy as the point was not raised by the appellants.

[^10]: I also note parenthetically that the trial judge added a descriptor to the standard that the designer must take all foreseeable risks into account no matter how unlikely or remote. Again, this wording does not, in my view, modify the Foundation standard in any way, which already requires the designer to take all foreseeable risks into account.

[^11]: I note that the trial judge's reasons, particularly at para. 76, contain an apparent inconsistency to the extent that the trial judge says both the risk was taken into account and the risk was not foreseeable. While I am aware that these two statements appear inconsistent, in my view the trial judge only said that the risk was not foreseeable in the sense that it was no longer foreseeable because the design had taken it into account. In any event, I do not believe this apparent inconsistency has any bearing on the result of this appeal, as the record and the trial judge's findings clearly support the conclusion he reached with respect to the application of the correct standard: the design took into account all foreseeable risks.