Steve,

I am trying to follow your reply. I agree with your position that full room involvement does not destroy fire patterns (they mostly evolve) specifically those on contents and wall/ceiling lining materials. However, I disagree that full room involvement does not obscure fire patterns on the floor. There are several studies that illustrate that full room involvement does obscure floor patterns - I have copied and pasted from an article that I helped with that describes this issue in more detail. In there you will see a summary of the full study that Doug posted. I have also posted the link to the article should you wish to review it in better formatting and look at the photographs.

Floor patterns
Fire patterns identified on the floor have been a common theme within fire investigation as being a possible indicator that flammable or combustible liquids were used within the fire (Smith 1983; Beyler 2009). In fact, a recent sentinel event analysis of wrongful convictions found that this one misconception is the most common factor in wrongful arson convictions (Bieber 2014). This misconception persists despite the warnings from both the fire science and fire investigation communities (Shanley et al. 1997; NFPA 2014; Gottuk and White 2008).

Many of the first texts on fire investigation discussed the concept of low burning and the importance of evaluating the floor for fire patterns (Kennedy 1959; Kirk 1969). In these texts the authors stressed that the investigator should evaluate low burns for possible ignition sources, but did not necessarily link the damage to ignitable liquids. In fact, Kirk was very adamant that investigators should not conclude that the damage was from an ignitable liquid as “such an interpretation was more often incorrect than otherwise” (Kirk 1969).

However, other texts of the time indicated that damage to floor was an indicator of arson (Battle and Weston 1960; Fitch and Porter 1968). Obviously this misconception was widespread as Kirk identified that it was “not uncommon for the investigator to assign the cause to the use of a flammable liquid” (Kirk 1969). More than a decade later this misconception can be seen in the majority of all fire investigation literature (Barracato 1979; DeHaan 1983; Smith 1983; Harmer et al. 1983; Kennedy and Kennedy 1985; Cooke and Ide 1985). The majority of these texts stated that the investigator should consider the damage to be caused by an ignitable liquid if the investigator would visibly observe damage to the floor in the shape of a puddle, have hard-edged burn marks in the shape of a pour, or the damage had the appearance of trailers (i.e. long lines of damage appearing to spread the fire from one location to another). However, most of these documents also cautioned against relying solely on the use of visible observations and encouraged the investigator to take samples of fire debris for analysis.

In the mid-1980’s there began a trend in the literature that spoke out against this misconception and began to provide a list of alternative explanations of damage to the floor (DeHaan 1983; Taylor 1985; Taylor 1986; DeHaan 1987; Eaton 1987; Wood et al. 2012). The studies demonstrated that the following causes could result in damage similar to irregular floor patterns, including: fires from interstitial space below the floor decking, melting plastics, draperies, furniture items, ventilation path and radiant heat from fully developed fires. NFPA 921’s original publication followed this trend and warned, “irregular, curved, or ‘pool shaped’ patterns on floors and floor coverings cannot always be reliably identified as resulting from ignitable liquids on the basis of observation alone” (NFPA 1992). Notice, however, none of these documents came out and directly stated that an investigator could not identify an ignitable liquid from a floor pattern based on observation, they only warned that it “cannot always be reliably identified” (NFPA 1992). This warning was strengthened over the years to say “irregular, curved, or ‘pool shaped’ patterns on floors and floor coverings should not be identified as resulting from ignitable liquids on the basis of observation of the shape alone” (NFPA 2001).

There have been a few studies performed that specifically evaluated the fire pattern creation on the floor (Putorti 2001; Mealy et al. 2013). Putorti (2001) performed a series of experiments that evaluated the damage to a variety of floor surfaces (carpet, wood and vinyl) with varying volumes of ignitable liquids used in the open. He evaluated gasoline and kerosene. He concluded that it was possible to identify the quantity of fuel used by the burn area. These tests were not conducted within a compartment. Mealy et al. (2013) conducted a series of compartment fire tests with ignitable liquids poured and evaluated the persistence of such a pattern through a compartment fire. They found that that floor patterns caused by ignitable liquids might be minimal because they can easily be destroyed and because the short duration of exposure due to fuel consumption.

Floor patterns were found lacking in many of the fire pattern tests where the compartment transitioned to a fully involved state (Shanley et al. 1997; Wood et al. 2012; Mealy et al. 2013). However, some data exists that indicates if a compartment fire does not transition to a fully involved state, then the floor patterns may persist (Putorti 2001; Mealy et al. 2013).

A study conducted in 2012 examined the effect of carpet underlayment/carpet pad on post-flashover fire, floor patterns (Wood et al. 2012). Specifically, the hypothesis that carpet pad seams could mimic the floor fire patterns previously attributed to ignitable liquid pours was examined. Fire tests in a scaled compartment using a propane sand-burner were designed to rapidly progress through flashover with a short period of full room involvement. Instrumentation included thermocouples in the gas layer and under the flooring material. Multiple carpet pads were tested. Carpet pad configuration was also varied including no seam and two, off-center seams for comparison and control purposes. Additional comparison and control samples were generated using ignitable liquid pours that achieved post-flashover conditions without use of the burner, but with the burner in place to maintain test consistency. A subset of replicate tests was also performed. Post-test data collection included examination, photography and a subset of depth of char measurements. Preliminary results indicated the ability to generate similar although not identical floor burn patterns between carpet pad seams and ignitable liquid pours (Figs. 7, 8 and 9).

Gorbett, G., Meacham, B., Wood, C., Dembsey, N. (2015). Use of Damage for Fire Investigations: a review of fire patterns analysis, research and future direction. Retrieved on April 22 2019 from: [link.springer.com]

Greg