Ancaster Fire Dept., Origin & Cause Inc., Activation Laboratories
* Dale Sutherland, Organics Manager, Activation Laboratories Ltd.
Ken Byers, Professional Engineer & Fire Investigator, Origin & Cause Inc.

On August 14, 1996, the Ancaster Fire Department conducted a training session to introduce new members to the proper approach for extinguishing a vehicle fire. Two vehicles, Car #1 (1982 Mercury Marquis Brougham, first fire set) and Car #2 (1985 Mazda GLC), were set ablaze at separate times. Through co-operation with the Ancaster Fire Dept., Origin & Cause and Activation Laboratories were able to carry out a study for the purposes of better understanding the performance of various materials that could be used as samples by an investigator. Knowledge of the best sample types may thus improve the detection of accelerants in vehicle fires. Specifically our aim was:

• to study the expected background of different intrinsic materials prior to pyrolysis
• to verify the sampling technique
• to investigate the retention of accelerants on different materials
• to study the preferred sampling sites
• to study the effects of time or weathering of the samples

Pre and post fire samples were taken from preferred locations in these vehicles based on the experience of the Origin and Cause Inc. investigator (from the investigation of hundreds of vehicle fires, many of them arsons, over a ten year period). When available, seat cushion foam, footwell carpet and underpad, and under seat carpet and underpad have provided the best samples for the analysis and determination of the presence of an accelerant if one has been used. Courtroom defense has sometimes suggested that oil, grease, or other repair shop fluids including kerosene and gasoline could be present at the footwell from use by service personnel or the vehicle owner tracking it in on the bottom of footwear. While only two vehicles were examined in this test, one objective was to determine if any background hydrocarbon material was present in these well used vehicles, and if this background would interfere with the detection and confirmation of the accelerant used, as may be claimed by defense council in court.

It is recognized that the very act of cutting seat and flooring material for background analysis from choice sampling locations can act to increase the possibility of the accelerant being found at, or near these locations after the fire as this alters the original surfaces and potentially the path that a splashed accelerant would take. Taking post fire samples elsewhere, distant from these locations, diminishes the credibility of the background analysis as being representative of the pre-fire conditions. In the real world of fire investigation we can only approach, not achieve, the ideal.

Eleven samples were taken prior to the exercise; ten samples were immediately taken after the extinguishing of the two fires; and seven samples were taken five days after the exercise. All samples were placed in one litre mason jars. As this was an initial study, it was not intended to be exhaustive. It is hoped that some of these observations will help to answer some of the “burning questions” that chemists and investigators may have regarding the fate of accelerants in vehicle fires.

Each sample was sealed in a one litre mason jar on-site. All tools were cleaned prior to, and in-between, sampling with a non-petroleum based Motomaster Citrus hand cleaner previously proven to not interfere with the identification of flammable liquids. The headspace above these samples was analyzed with a passive sampling technique (modified ASTM E1412-91). As per the laboratory’s standard procedure, each sample was analyzed by Gas Chromatography coupled with Mass Spectrometry (GC/MS, ASTM E1618-94). The data was compared to a library of flammable liquids by way of the review of 105 reconstructed mass ion profiles specifically designed to differentiate between flammable liquids and possible pyrolysis products.

Prior to the fire, seven samples were taken to obtain information on their natural background using the above method of analysis. Each sample was taken from the driver’s area of the car unless otherwise noted. Both vehicles exhibited similar background levels for the various samples. The following is the list of materials tested and their results, in order of increasing background, thus from most preferable to least preferable sample type.

1. Chip foam: coarser blue and white chip foam used as a second layer in some car seats, (Car #1)
2. Footwell underpad, (Car #2)
3. Footwell Carpet (Both Cars) and Carpet Mat, (Car #2)
4. Seat foam and material, (Both Cars)
5. Sound deadening pad, fibre based (Car #1)
6. Underpad under driver’s seat, (Car #2)
7. Carpet under driver’s seat, (Both Cars)

The chemical analysis of the carpet under the driver’s seat was significantly different from the response for the material at other locations and was substantially greater than a laboratory procedure blank. Carpet in this location would be a poor choice for a comparison sample or for analysis of a post-burn sample containing a low amount or highly weathered flammable liquid. The complexity of the chemical background was significantly worse for Car #1 but was still considered as poor quality for Car #2. Although of poor quality, since GC/MS was used to filter out the sample background by monitoring mass ions specific to flammable liquids there was no presence detected of any flammable liquids in these pre-burn samples.

Burn #1:

Approximately 1 Litre (1 quart) of gasoline was carefully applied to the front seat cushion in Car #1 and ignited. A significant quantity of seat foam survived the fire even though extensive fire damage occurred in the passenger compartment.

• The yellow foam from the seat surface retained five times more gasoline than the underlying blue/white chip foam.
• The responses from both foam samples were very strong and essentially indistinguishable from fresh, un-weathered, gasoline.
• Blackened window glass that had fallen into the drivers footwell was negative.

Burn #2:

Approximately 2 Litres (0.5 gallons) of a 50:50 mix of lamp oil and fuel oil was used to ignite the front and rear seat cushions of Car #2. Very little seat foam survived. Floor carpeting was thoroughly covered by fire debris but was not extensively damaged.

• Remnants of foam from the driver’s seat tested positive and was easily the strongest response of all post burn samples.
• Carpet from under the driver’s seat tested positive and significantly stronger than the carpet from the driver side footwell.
• Blackened window glass that remained intact on the passenger rear door provided a positive result of significant strength.
• A puddle of water used to extinguish the fire was found in the front passenger footwell. A paper towel was used to sample this puddle by letting it float on the water for 3 to 4 seconds. This sample was positive but was half of the strength of the blackened window glass.
• Carpet from the driver side footwell was the weakest positive response that was obtained.

Second Visit – Five Days After Burn:
After 5 days of relatively hot summer weather (> 25°C or 77°F), additional samples were taken from both vehicles.

There were four samples taken from the Car #1:

• The driver’s seat chip foam was less weathered. It had a similarly strong response and was easily identified as a positive.
• The driver’s seat foam now had a moderately weathered result for gasoline with a sufficiently strong response for easy identification as a positive.
• The sample of carpet mat from the footwell was also very weak. Some highly weathered traces of gasoline could still be identified but in this case, this sample would not have been reported as a positive by GC/MS due to the overall low response. This sample would have been a candidate for reanalysis by GC/MS/MS. Using methodology developed in April, 1997 this sample was reanalyzed by GC/MS/MS in January, 1998, 17 months after sampling, and found to be positive for the presence of gasoline.
• The sample from the footwell underpad was very weak. Some highly weathered traces of gasoline could still be identified but also would not have been reported as a positive by GC/MS due to the overall low response. This sample would also have been a candidate for reanalysis by GC/MS/MS. Reanalysis by GC/MS/MS in January, 1998, was negative.

There were two samples taken from Car #2:

• The analysis of the driver’s side footwell underpad showed a strong response for the Medium Petroleum Distillate (MPD) mixture used to accelerated the fire.
• The sample from the driver side footwell loose carpet mat was still stronger for the Medium Petroleum Distillate (MPD) mixture and was less weathered than the underpad.

• Seat cushion chip foam is an excellent material to sample if it can be found after a fire.
• Any pieces of foam are preferred samples, no matter how small.
• Carpet taken from beneath the driver’s seat is not a good comparison sample, but may be a good sample after a fire, especially if it’s the only one available.
• Residues of soot found on glass is carbon, which is an excellent absorber of volatile compounds if it is of the soft porous type. In this study we
• found that a relatively thick sample of soot could provide a positive result.
• The surface of standing water in the vehicle can be a useful sample.
• Different materials have different backgrounds and absorb differently. If possible, separate these sample types as composite samples significantly add to the complexity of the analysis.
• The use of Motomaster Citrus Hand Cleaner as a cleansing agent for tools is justified but care should be taken to use it sparingly and to ensure that surfaces are wiped thoroughly to remove the cleaner prior to taking the next sample. Excess residue from the cleaner will not interfere in the identification of a flammable liquid but can complicate the analysis and would require clarification in court. It is suggested that if Sterno is suspected to be present, the hand cleaner should not be used and water should be substituted. Under very cold conditions this hand cleaner will freeze.
• When using a paper towel for collection of hydrocarbon residues from the surface of water, or for wiping soot covered glass, it is best to use a new plastic wrapped roll of paper towel rather than an opened and previously used roll.
• Although the analysis of samples taken five days after the fire still resulted in positive identification of the flammable liquids used as accelerants, a significant amount of the flammable liquid had been lost due to weathering. It is the experience of this laboratory that some flammable liquids can still be detected even months after the event.

This is only one study, and as such can only provide further insight into the techniques and types of samples most useful for fire investigators to take, that are more apt to result in more definitive laboratory analyses. It cannot be considered as comprehensive since other materials, accelerants and environmental factors may be different in other cases.

The study convincing demonstrated the value of laboratory analysis by GC/MS. The specificity of this technique combined with the highly sensitive full scan operation of an Ion Trap Mass Spectrometer was essential at confirming the presence of an accelerant in the samples taken five days after these test burns. The confirmation analysis by GC/MS/MS is an important new technology for analysis of highly weathered samples. Here it has demonstrated the ability to report a positive result for a sample having such small amounts of a flammable liquid that it would have been reported as negative by GC/MS, even though the sample was 17 months old.

Primary Author

Mr. Dale A. Sutherland
Activation Laboratories Ltd.

Mr. Sutherland is a Forensic Scientist and holds the position of Organics Manager and Director of Research at the Activation Laboratories Ltd. head laboratory in Ancaster, Ontario, Canada. This laboratory is 1 of 18 facilities that cover Canada, the United States, Mexico, Brazil, Chile, Peru, Bolivia, Argentina and Australia. Mr. Sutherland has three University Degrees and has obtained the distinction of Charted Chemist with the Association of the Chemical Profession of Ontario. In his 19 year career, 17 years has been devoted to chemical analysis using Gas Chromatography / Mass Spectrometry (GC/MS) which has resulted in the development of over 22 publications. In past positions he received recognition as a Research Scientist with the Ontario Ministry of the Environment, an Award for Innovation from a large engineering firm and became Director of Dioxin Technologies at a private laboratory. At Activation Laboratories he has turned his expertise and resources to the investigation of the chemical contents of Fire Debris using the latest in GC/MS and new GC/MS/MS technologies. He is a member of the Canadian Society of Forensic Sciences Chemistry Section Sub-Working Group. Mr. Sutherland has studied and endeavored to develop services specifically designed for the Arson Investigator and for presentation in the court system and has been qualified in court as an expert witness.

Associate Author

Mr. Ken D. Byers, P.Eng., CFEI
Origin and Cause Incorporated
Phone: 905.648.5522
Fax: 905.648.4101

Mr. Byers is a Professional Engineer (Mechanical) and is part-owner of Origin and Cause Inc., a consulting forensic engineering firm based in Ancaster, Ontario, Canada. He has been investigating fires and mechanical failures for the past twelve years. Concentrating on vehicle fires, he has carried out approximately twelve hundred investigations for many insurance companies as well as vehicle manufacturers. He has been accepted in court as an expert witness many times.

Mr. Byers has taught Vehicle Fire Investigation at IAAI Chapter Seminars (he has been a member of the Ontario Chapter since 1986) and lectured at Insurance, Fire Marshall and Investigator Training seminars. He was a part-time fire-fighter with the Ancaster Fire Department at the time of the two vehicle test burn and has had a special interest in the detection of accelerants in fire debris.