Determining Arson in an Origin and Cause Report

The NFPA 921 guideline for fire and explosion investigations acts as a peer-reviewed, consensus document that addresses the science of fire investigation and the methodology for performing such investigations. Updated approximately every three years since its inception in 1992, this guide not only provides a basic source of knowledge of fire science for fire investigators, it also becomes a tool for challenging the qualifications of those fire investigators and their expert opinions in court.

While the process of elimination is “an integral part of the scientific method” involving “elimination of a testable hypothesis by disproving the hypothesis with reliable evidence”, NFPA 921 maintains that negative corpus is an inappropriate use of the process of elimination because it involves “identifying the ignition source of a fire by believing to have eliminated all ignition sources found, known, or suspected to have been present in the area of origin, and for which no supporting evidence exists.”2

In the case of Heiser’s Health and Fitness Ltd. v. Saskatchewan Government Insurance 2019 SKQB 124 [Heiser’s], a fire broke out in a gym about a month after a flood rendered the building out of commission. The fire further destroyed the building beyond repair. The cause and origin investigator had ruled out various possible causes of the fire3 and, when left with no other possible causes, concluded that the fire was incendiary, despite the absence of any physical evidence to support his conclusion, a fact of which he even acknowledged in his report. His report was released in early 2011, just months before the 2011 edition of NFPA 921 was released, in which it was stated for the first time that the process of negative corpus was inconsistent with scientific method.

The expert in question defended his approach in cross-examination saying “the absence of physical evidence may be physical evidence”, which the court aptly noted appeared “inconsistent with both logic and the overall tenor of the scientific method espoused in the Guides.” In rejecting the expert’s opinion, the court found that the “conclusion was reached by default, after a process of elimination of other possible causes, instead of being built upon a foundation of existing evidence.” In reaching this decision, the court further opined that the opposing experts’ criticism of negative corpus “accord[ed] with the clear wording of the investigative Guides since 2011”, but also “follow[ed] the spirit and intent of the Guides.”

The duty of the expert is to evaluate the evidence independently. While in some cases, it might be frustrating to be in the position where no other conclusion than “undetermined” for the cause of the fire can be justified under the NFPA 921 approach, the expert can take comfort in the fact that the cause and origin investigation is not the final word in whether a court can find that a fire was arsonous. To this point, Mr. Justice Tochor explains in Heiser’s that:

A finding that the cause of the fire is undetermined does not automatically prevent a finding of arson. Nor is it essential to conclude the fire was incendiary before going on to assess whether arson occurred.

In cases of suspected arson, the courts recognize that the physical evidence of arson will often be destroyed in the fire. Accordingly, the courts use a holistic approach to consider the totality of evidence, which includes not only the fire investigation evidence, but also other evidence relating to motive, opportunity, financial difficulty and other inculpatory circumstances which link the accused to the fire.4 In short, the insurer can lead evidence of a suspicious nature that implicates, but does not conclusively prove on its own, the plaintiff of deliberately starting the fire. Fire investigators will often collect such evidence as a part of their fact finding process, but should not allow such evidence to colour their investigation into the cause of the fire.

This case echoes the findings of the British Columbia Supreme Court in Number 216 Holdings Ltd. v Intact Insurance Company 2013 BCSC 1267, a case which also dealt with an initial report upon which the court could not rely due to the investigator’s reliance on negative corpus. Indeed, Mr. Justice Tochor, in reaching his decision, commented that he was “in the same position as the trial judge in Number 216 Holdings Ltd. v Intact Insurance Company”, where, “despite finding some level of suspicion as to the plaintiff’s involvement, [the trial judge] concluded the balance of probabilities had not been tipped in favor of the insurer.”

From a review of these cases, cause and origin investigators, as well as those adjusting such fire loss claims, should be alive to the issue of needing to revisit any investigations that relied on negative corpus. This might involve preparing a supplemental report, but, at a minimum, should involve acknowledging that the standards of NFPA 921 have changed and, as such, would oblige the investigator to revisit their investigation. Such acknowledgement is not fatal to the insurer’s claim that a fire was deliberately set.

1NFPA 921, 2017 ed., subsec. 19.6.5. The definition of negative corpus in the NFPA 921, 2011, subsec. 16.2.5, which is the edition discussed in the case addressed in this article, is substantially the same.
3The primary sources the investigator considered were: an electrical failure, a vapour barrier, a dryer in the laundry room and papers from the storage room. However, the investigator failed to consider other sources: spontaneous combustion of the wet wood chips used as insulation in the roof, the dryer’s vent (in light of the plaintiff drying towels prior to the fire), and evidence of a person being shocked after touching an electrical component.
4See, for instance, Rizzo v. Hanover Insurance Co. (1993) 14 OR (3d) 98 (ONCA), Bezdziecki v. Allstate Insurance Co. of Canada [1998] OJ No 4853 (ONCA), Savoie v. General Accident Assurance Company (1999) 210 NBR (2d) 74 (NBCA), Lancer Enterprises Ltd. v. Saskatchewan Government Insurance 2011 SKCA 28 and Richardson v. Smith and State Farm Mutual Automobile Insurance Company 2012 NBCA 75.