Determining the Fire’s Point of Origin: Beyond the Obvious

THE AVERAGE FIRE INVESTIGATOR WILL INVESTIGATE at a minimum hundreds of fires during his career. Some will be extremely extensive operations involving numerous outside agencies and restraints. Others will be single-engine responses with little or no damage involved. The fact is that the extent of the fire often depends on when the fire is discovered and the actions taken when it is discovered. Unless someone uses an accelerant or sets up a scenario for a rapid spread of flames, all fires start at one small point. Sprinkler systems not only greatly reduce the damage of a fire but also help to preserve the initial area of origin. Generally, we are sometimes unable to locate the exact point of origin, which further limits our ability to give a definite cause of the fire. If we enter into an investigation where the scene is untouched, we often can determine the fire’s origin and cause. The greatest ally we have to make this possible is an engine crew or command that allows the investigation to begin before extensive overhaul is underway. Leaving the scene as it is found is paramount in many investigations. Occasionally, we encounter the extreme condition.


Recently, our department responded to such a call. After arriving on the scene, the first-in engine company found the sprinkler system had put out the fire. The crew decided to take no further action until the investigator was able to evaluate the scene. The fire was out because the sprinkler system extinguished the fire in its incipient stage, as designed. This left the room in the same condition as when the fire first started.

The fire investigator arrived and conducted his investigation according to NFPA 921, Guide for Fire and Explosion Investigations. Using the systematic approach, he finally arrived at the room of origin and began to rule out items. When he had eliminated everything in the room, only one possible cause of the fire was left: an intentional human act.

By that time, he had limited the possible suspects to two of the occupants of the house. Both were interviewed; both denied having any knowledge of how the fire started. Again, doing his job, he turned the investigation into an accusatory mode but still was unable to produce information to complete the cycle needed for determining the cause of the fire. Throughout the investigation, photographs were taken and the scene was well documented.


Our department uses peer review whenever possible and always when a question arises on an investigation. While reviewing the report, I found that the scene was well documented and the investigation was properly conducted. I was able to confirm that the two occupants had the means and opportunity to have set the fire, but neither had any motive of which we were aware.

While reviewing the photographs, I realized they were taken about an hour after the fire had taken place. The sprinkler system had extinguished the fire in its incipient stage; this left the perfect stage for viewing photos (photo 1). While viewing this photo, I felt that the shadow in the sink may have been farther to the left during the time of the fire, and I noticed that the mirror on the counter sat directly below where the initial flame was ignited.

(1) The fire scene. (Photo by Certified Fire Investigator Jason Garner.)

That’s when I first started to develop the theory that the mirror may have been responsible for the fire. If the mirror was a magnification cosmetic mirror and if the sun traveled across its path, the mirror would have to be eliminated as a possible cause. I requested another certified fire investigator to assist me in testing my theory at the residence. After gathering the equipment needed to conduct the test, we contacted the resident and requested a second evaluation of the fire scene. The resident agreed and was happy to assist in any way possible. We arrived at the house approximately 20 minutes before the time of the initial dispatch, and weather conditions were very similar. At the time of the fire, the weather was as follows: 39 degrees, dew point 12.2, humidity 34 percent, visibility 7.0 miles, and clear skies.


Entering the house, we were taken directly to the room of origin. I did not repeat the typical systematic approach, since the fire had already been investigated and the room of origin was obvious. First, I requested that the room be returned to the exact condition it was prior to the fire. Initially, the resident said she was told to leave the room as it was by the previous investigator and did not want to move anything. I explained my role and position in the investigation, and the room was set up as it had been before the fire. I did not let the resident know what we were doing other than to say we were taking a second look at the fire.

The first thing I found was that the sprinkler that had extinguished the fire sat directly above the door in the position she had left it in. This immediately extinguished the fire in its incipient stage. Prior to moving the mirror to reflect the sun’s rays, it was obvious that the sunshine was coming directly in the window and hitting the mirror, as I had predicted. The bag seen in photo 2 was found on the floor. I hung it on the door. The resident stated that the bag had been hung much lower. This, once again, perfectly lined up the origin of the fire on the bag and the door. The bag was used to hold hair-care items such as hair ties and brushes. Photo 2 represents the placement of the mirror in comparison with the photo from the fire scene investigation (photo 1). The reflection on the door was, again, as expected.

(2) Testing the theory. (Photo by author.)

Using our hands, we could feel the heat generated but were unable to produce enough heat to start a fire. While conducting the test, we made a full photo and video documentation of the re-creation. At this point, I felt the scene was exactly as it was the day before the fire started. The normal temperature in the room was elevated from the direct sunlight entering the window. This was another indication that nothing obstructed the rays of the sun from penetrating the room. Using a digital laser thermometer, we were able to capture elevated temperatures but were unable to get enough heat generated to start a fire. The test continued for about 45 minutes.

As luck would have it, I found that the actual distance for the concentrated beam of light was about six inches farther from the mirror than the area I had originally thought was the hot spot. While trying to reposition the door and mirror, my right arm was hit by an immediate pain that made me move rapidly. Returning to the same point and slowly moving my hand toward the mirror once again, I found that the intense heat was being reflected. The entire test was being conducted six inches too close to the mirror.

As photo 1 indicates, the bag was hung where it was to attempt to set it on fire. What I failed to realize or consider was that the contents of the bag were destroyed in the original fire, and this was not re-created. Also, the bag would have been hanging lower on the door. After moving the bag down to the correct position and opening the door a few more inches, everything came together. Since time had passed and the sun was moving toward the sink, I held the mirror and tried to reproduce the concentrated beam of heated light. It did not start a fire, but we then knew the possibility of a fire did exist.

After repeated attempts, I finally found the problem: The mirror had to be perfectly still to generate enough heat to burn holes in paper. By holding the mirror with my hands, the slight movement of the point where the beam contacted the paper did not allow the temperature to build high enough to ignite the paper used in the test. Moving a quarter inch closer or farther away from the paper would not generate the heat needed to ignite the paper either. I set the mirror at a stable point and established the exact distance from the mirror to the paper and was able to burn holes through the paper at will. When the relationship of the correct distance, angle, and stability was understood, there was no doubt that the mirror was responsible for the fire and that the fire was accidental-caused by the reflection of the sun’s rays.

. . .

This fire would have been considered a probable arson if the sprinkler system had not preserved the scene in its incipient condition, and it would have taken numerous hours of fire department investigation and interrogations to try to determine the responsible party. There were three occupants in the residence at the time of the fire. One was an elderly female who was nearly confined to bed; the second was the 57-year-old homeowner; the third was a 15-year-old female. Taking into consideration the elderly female’s condition and the location of her room relative to the area of fire origin, it was highly possible that a fire would have resulted in a fatality had it not been for the installed sprinkler system.

I have attempted to research to find other fires that started in a similar manner and found the scientific proof to be very weak. As mentioned, normally the scene is not found in a condition that allows the investigator to determine exactly how the scene was prior to the fire. A thorough investigation is a must in every fire, and there is only one chance to get extensive photo documentation. By using peer review and revisiting the scene through notes and photos, we often find items we may have overlooked or originally neglected to connect to the fire scene.

Just because a fire is small does not mean it was an accident, and not all large fires are arsons, of course. Approach every fire with the same systematic approach, and follow your guidelines to ensure every effort is made to come to a solid conclusion. Remain open-minded to the unexpected. The odd or unexpected is what generally brings you to the situation that caused the fire.

JESS ZERBE, certified fire investigator, is a lieutenant in the Marietta (GA) Fire Department, where he heads the fire investigations branch of fire prevention.


No posts to display