Established in 1990

Publications

THE EFFECTS OF WATER SUBMERSION ON THE DETECTION OF FLAMMABLE LIQUIDS

Robert Eyford, CFEI, CCFI, CFII, CVFI
Rae-Tech Investigations Ltd.

I recently had occasion to use our Canine Fire Investigation Team in a fire scene where the team located flammable liquids in several locations. This particular fire scene had been tied up by public fire investigators and the police for a period of two weeks prior to us being able to enter the scene. During that period of time the fire scene had been pretty well continuously drenched with coastal rain. This was a residential structure where the roof had totally burned away, allowing rain to continuously soak the interior floors. We continued to work the fire scene for about ten days, including searches for flammable liquid odours by the canine Ina and during which time the rains continued nearly non-stop. Discussion surfaced on the fact that Ina could still detect the odour of flammable liquids after several weeks of continuous rain, which in turn raised the question of how effective a canine might be in the detection of flammable liquid odours that had been totally submerged in water for a period of time, for example a marine vessel which had sunk following a fire. Casual inquiries within the industry revealed that it is generally felt that once a burned vessel had been submerged for a short period of time, there would be very little possibility of obtaining evidence of flammable liquids that had been poured inside the vessel immediately prior to the fire. Thus arose the need for this experiment.

For the purposes of this experiment, four sets of four pieces of 2" X 2" wood, 18 " long were prepared. These were divided into the following categories:

Category A in which all pieces of wood were burned and Category B in which none of the pieces of wood were burned. These categories were further divided into two groups. Category A was divided into groups 1 and 2, which were submerged and not submerged respectively. Category B was also divided into groups - 3 and 4 which were also submerged and not submerged respectively as shown in the chart below.

Category A
(Burned)
Category B
(Unburned)
GROUP 1
(SUBMERGED)

4 pieces of wood with 1 exposed to flammable liquid.
All pieces burned
on one end. 
GROUP 2
(DRY)

4 pieces of wood with 1 exposed to flammable liquid.
All pieces burned
at one end. 
GROUP 3
(SUBMERGED)

4 pieces of wood with 1 exposed to
flammable liquid. 
GROUP 4
(DRY)

4 pieces of wood with 1 exposed to
flammable liquid. 


One piece of wood from each group was exposed to flammable liquids by having a four inch section of one end submerged into a container of flammable liquids for a period of one minute. The flammable liquid was fresh gasoline mixed in a 50:1 ration with 2 stroke engine oil. The mixture had earlier been prepared for use in a chain saw. Both exposed pieces of wood from Category A (burned) were then set on fire with a propane torch and allowed to burn freely until combustion no longer continued. All of the non exposed pieces of wood from Category A were burned on one end in like manner, even though only one piece of wood had been exposed to flammable liquid. The exposed pieces of wood were kept separate from all other pieces of wood in their group as well as separate from each other to avoid cross contamination.
 
The experiment began on July 30, 2001 and took place in a small creek in a remote corner of the writer's property. The creek was about six feet in width and the water generally flowed at approximately five miles per hour. The creek was free of pollutants and was several miles away from the nearest human habitation. A fast moving creek was chosen as compared to still water so that flammable liquids leeching away from the exposed pieces of wood would be taken downstream and not allowed to re-contaminate the pieces of wood.

From Category A (burned) the two pieces of wood that had been exposed to flammable liquid were placed in their respective positions. The exposed wood from group 1 was placed in the creek and allowed to float freely while attached to the shore with a nylon rope. The exposed wood from group 2 was placed in a clean five gallon plastic bucket and stored in a dry place nearby. The unexposed pieces of wood from group 1 were also placed in the water, but upstream from the exposed piece of wood. They were fastened to the shore with nylon rope.

From Category B (unburned) the exposed piece of wood from group 3 was submerged with a large stone on the opposite side of the creek from the wood from group 1 but without covering the end that had been exposed to flammable liquid. This piece of wood was also anchored to the shore by a nylon rope in case it became dislodged from the stone. The exposed wood from group 4 was stored in a separate, clean five gallon bucket by itself so it could remain dry throughout the experiment. The unexposed pieces of wood from group 4 were placed in a separate dry five gallon bucket and kept dry.

The following day, August 1, 2001, the accelerant detection canine Ina was taken to the experiment area by the creek. All of the pieces of wood from groups 1 and 2 were placed in two separate lineups. The two exposed pieces of wood, one submerged and one not submerged were placed at random in the two lineups. For the purposes of this experiment, the lineups were moved to a different location a few feet apart between searches. Each lineup was searched twice in different order by Ina. When directed to search these lineups, Ina successfully picked the exposed piece of wood from the lineups each time.

The same sequence took place with groups 3 and 4 and again Ina was able to pick the exposed piece of wood out of each lineup. Following the searches, the exposed and submerged pieces of wood were replaced in their same positions in the creek and the exposed but not submerged pieces of wood were replaced in their respective dry buckets.

The searches continued every other day in the same manner for the first two weeks with the same results. After the second week, the search schedule was changed to once per week with no change in the results. After two months had passed, the searches were conducted every two weeks until three months had passed. The results were the same after three months. No matter where in the lineups the exposed pieces of wood were placed, Ina was still able to find the pieces that had been exposed to the flammable liquid.

At the end of the three month period the experiment was stopped. We concluded that the accelerant detection canine Ina was able to detect the odour of flammable liquids on a piece of exposed wood that had been submerged in water for three months as easily as a similar piece of wood that had not been submerged in water. We also concluded that it does not make a difference whether or not the piece of wood has been burned following exposure to the flammable liquids.