The Essential Facts

This is not a story of slippery or drunken marine mammals. It is an insurance story.

In the early 1990s, the Canadian National Railway Company ("CNR") embarked on a project to have a tunnel boring machine ("TBM") designed and constructed to run under the St. Clair River to connect Sarnia, Ontario and Port Huron, Michigan. The TBM was a massive machine. One of the key design considerations was keeping dirt and debris out of the bearing housings of the TBM. Apparently, dirt in the bearing chamber is a very bad thing. To this end, the TBM design team integrated twenty-six specially lubricated seals into the TBM to act as a "gauntlet" to protect the main bearing from dirt.

CNR purchased an "all risks" policy of insurance (the "Policy") from a number of insurers (the "Insurers") and paid a hefty premium for same. The Policy 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" but excluded "the cost of making good…faulty or improper design" and "inherent vice".

As the tunnel was being excavated by the TBM, engineers discovered that some amount of dirt had gotten into the main bearing chamber. Construction was halted, the TBM was repaired, modifications to the seals made, and the tunnel completed. The delay in completing the tunnel was a costly one; CNR claimed on the Policy, the Insurers denied coverage, and CNR sued.

The evidence at trial established that the probable cause of dirt making its way to the sacred bearing chamber was the deflection (movement/bending under stress) of metals in the TBM beyond what the engineers had anticipated in the design. The engineers had anticipated and accounted for some deflection of metals in the design but the degree of deflection that actually occurred was much greater.

The Trial Decision

The trial judgment can be read at http://www.canlii.org/eliisa/highlight.do?language=en&searchTitle= Ontario&path=/en/on/onsc/doc/2004/2004canlii33029/2004canlii33029.html.

At trial, Justice Ground awarded CNR more than $30,000,000, including costs. The trial judge found that the design and construction of the TBM had taken all foreseeable risks into account and that the design of the TBM was not "faulty" or "improper", disentitling the Insurers from relying on the exclusion in the Policy. The judge accepted that deflection of metals will occur under stress but found that the degree of deflection that actually occurred was beyond what was foreseeable when the TBM was designed. As the designers had considered foreseeable risks, designing the TBM to a very high level, the trial judge held that the design was neither "faulty" nor "improper". The loss was covered and the Insurers had not discharged their burden of establishing that the exclusion applied.

The Court of Appeal

The decision of the Ontario Court of Appeal can be read at http://www.canlii.org/eliisa/highlight.do?language=en&searchTitle= Ontario&path=/en/on/onca/doc/2007/2007onca209/2007onca209.html

The Ontario Court of Appeal, in a 2-1 split, allowed the Insurer's appeal. The essential reasoning of the majority was that the very fact that the TBM failed is de facto proof that there was a faulty or improper design, triggering the exclusion in the Policy.

The Supreme Court of Canada

The decision of the Supreme Court of Canada can be read at http://www.canlii.org/en/ca/scc/doc/2008/2008scc66/2008scc66.html.

In a narrow majority (4-3), the Supreme Court of Canada allowed CNR's appeal and restored the award of the trial judge.

The Insurers' argument was that where the failure occurs under conditions that were foreseeable, the combination of that failure and the foreseeability of the element that lead to the damage should be sufficient to trigger the exclusion.

CNR's argument was that the TBM need not be designed to a standard of perfection. Further, because the TBM was designed and constructed to the state of the art (essentially to the best knowledge and practices of the day), the design was neither "faulty" nor "improper" so as to trigger the exclusion.

In finding for CNR, the majority made the following observations:

  • When considering whether a design is "faulty", the court must focus on the design itself - not the designer's fault or lack thereof [para 41];
  • To rely on an exclusion containing the words "faulty or improper" in relation to the design, an insurer must establish that the design fell below a "realistic" standard [para 53];
  • A design is not faulty or improper simply because it does not meet a standard of perfection in relation to all foreseeable risks [para 51];
  • While designing and constructing the TBM to the state of the art was sufficient to maintain coverage and not trigger the exclusion, the outcome may have been different had the TBM been designed and constructed to the lower "industry standard" [para 54]; and
  • It was open to the insurers to write the exclusion to exclude coverage for losses arising from "design failure" or "design failure in conditions of foreseeable risk" - suggesting that had the exclusion been so cast, the outcome of the case may well have been different [para 56].  

The distilled message conveyed by the majority decision is that when insuring something that has never been built before (a novel design), it will be covered if a failure occurs despite the exclusion for "faulty or improper design" if it was designed and built to the state of the art at the time.

The essence of the minority decision is if the item insured fails as a result of an inability to withstand or cope with risks that were foreseeable, its design is "faulty" and the exclusion should apply. The minority decision makes the interesting observation that the majority decision effectively turns the Policy into a warranty.

A full scale discussion of the significance of the decision is beyond the scope of this article. Given the close divide in both the Court of Appeal and the Supreme Court of Canada, this case will almost certainly not be the final word on this particular exclusion clause. A very different outcome may well be possible on only modestly different facts.