A number of recent cases illustrate the exacting standard applied by the court to questions of competing-causes in fire claims.
Where, who and how
The most recent case is United Marine Aggregates Ltd v GM Welding & Engineering Ltd, which shows that working out where the fire started – or who started it – may not be enough. If you are to succeed in court, you also need to know how it started.
In United Marine, the claimant’s processing plant was badly damaged by fire. The claimant and the defendant had both carried out ‘hot works’ in separate parts of the building shortly before the fire broke out. Therefore the first task of the judge (Edwards-Stuart J) was to work out where, roughly, the fire had started. In other words, which party’s hot works caused the fire.
After hearing expert evidence from three fire experts, the judge decided that the fire had been caused by the defendant’s oxy-propane cutting works. The investigation did not end there, however. The claimant had to show that the fire was caused by the defendant’s negligence, and this meant showing exactly how the fire was caused.
Edwards-Stuart J identified eight possible mechanisms, depending on the various different directions a globule of molten steel might travel. Having identified which, on the balance of probability, was the actual method of ignition, the judge had to decide whether or not the defendant was liable.
Unsurprisingly, the claimant relied on the doctrine res ipsa loquitor: if the defendant had carried out its task with reasonable skill and care, the fire would surely not have occurred. Edwards-Stuart J held that the maxim did not apply. The claimant had to show that:
- The defendant had not shown reasonable care.
- The lack of reasonable care had caused the fire.
In fact, the precise mechanism of the fire starting was not reasonably foreseeable. The defendant was therefore not negligent for failing to prevent it.
This judgment illustrates in stark terms the need for the court to be satisfied precisely how the fire occurred, and not simply where it started and when. In most cases this requires evidence from expert witnesses. These will depend heavily on a site investigation, and any documentary and witness evidence that may be available.
What was the exact cause?
Unfortunately, proving the exact cause and mechanism of a fire is a common problem, particularly where the extent of the fire damage means that there are a number of potential competing causes.
In Amsprop v ITW Ltd, for example, considerable damage was caused to the Hard Rock Cafe on Piccadilly when a fire broke out in the kitchen. The fire experts managed to agree that the fire started when grease in a flue vent above a cheese melter (known as a Salamander) caught fire. The defendant engineering company had carried out a service and repair to the Salamander two days before the fire.
The question of liability turned on whether the fire was caused by a flame from the burner on the Salamander (in which case the defendant might be liable) or a flare up from food on the grill (in which case the defendant was not liable).
The claim failed. HHJ Toulmin CMG QC could not conclude, on the balance of probability, which of the two potential scenarios had actually occurred, and suggested it would be mere “speculation” to do otherwise (paragraph 105, judgment).
The Sherlock Holmes fallacy
In Amsprop, the court considered the application of the Sherlock Holmes fallacy, discussed by the House of Lords in Rhesa Shipping Co SA v Edmunds (The Popi M) [1985] 1 WLR 948. The Sherlock Holmes fallacy is the doctrine that, when you have eliminated the impossible, what remains, however improbable must be the truth. In a legal context, if none of the options are established on the balance of probabilities, the claimant will not have proved its case.
In Fosse Motor Engineers Ltd v Conde Nast & National Magazine Distributors Ltd, for example, the claimant’s warehouse was destroyed in a fire. Expert forensic engineering evidence could not identify the cause of the fire. The court therefore had to consider five possible causes of the fire, one of which was that the defendant’s agency staff had carelessly discarded a cigarette. Akenhead J said that the “explanation is anything but ‘elementary'”, and the cause of the fire could not be proven on the balance of probabilities (paragraph 107, judgment). Accordingly the claim failed.
By contrast, in Milton Keynes Borough Council v Michael Nulty, Edwards-Stuart J held that there were only three possible causes of a substantial fire to a recycling centre. While none of them were inherently likely, two of the possible causes were “very much less likely” than the third. On that basis the third was the probable cause of the fire. This was a decision that he could only reach after the expert evidence had concluded that there could only be three causes of the fire, and two of them were very much less likely. In many fire cases, such an analysis will not be available, and the claimant may not succeed.
These cases all show the difficulties inherent in proving causation in fire cases where much of the evidence has been destroyed. They serve as a reminder that there is no substitute to a detailed forensic report from an appropriately qualified fire investigation expert, and early compilation of all available evidence.
