by Tony Cafe

First printed in Firepoint Magazine December 1993, Australian Fire Investigators Association.


Recently some members of the NSW Chapter of the IAAI witnessed demonstrations of aids which could be used to assist an investigator at the fire scene select samples for later laboratory analysis. A sniffer dog trained to detect accelerants was demonstrated, as well as a portable gas chromatograph.

The use of various aids and techniques to detect accelerants at fire scenes has attracted controversy during the previous ten years. The much maligned sniffer has suffered continual criticism, yet the author has found it to be invaluable on some occasions. The use of physical indicators such as floor burn through to indicate the presence of accelerants continues to be debated in court. Experts using this sort of evidence without the support of positive laboratory evidence are often heard resorting to hyperbole and analogy to support their case, which usually ends in indignation. The fact is, these techniques are solely used as aids to detect the presence of accelerants, and samples should be submitted to the laboratory for confirmation of the presence and identification of the accelerant.

The aim of this article is to discuss the properties of accelerants and to give an overview and evaluation of the various techniques which can be used to assist the investigator sample debris at the fire scene. As the topic is regularly debated at seminars, views opposite to those in this article will inevitably surface. I'm sure readers would enjoy hearing these views in future issues of "Firepoint"

The Common Accelerants

The most commonly used accelerants are petrol, kerosene, mineral turps and diesel. These accelerants are generally complex mixtures of hydrocarbon molecules, These hydrocarbons have similar chemical properties, however their boiling points vary and cover a wide range of values. This variation causes the accelerants to alter their composition during the evaporation process. The more volatile hydrocarbons evaporate at a faster rate leaving the heavier hydrocarbons in the debris and after a period of time the accelerant becomes less volatile and less abundant.

During the evaporation process, the headspace above the accelerant becomes concentrated with the more volatile hydrocarbons and so has a different composition than the accelerant left in the debris, It is this headspace which Is tested by the various techniques to detect the presence of an accelerant amongst the debris.

Heavy accelerants such as diesel, or accelerant residues which are heavily evaporated will be difficult to detect as they provide little vapour in the headspace. Accelerants trapped under compacted soil and debris will also be difficult to detect so the debris must be disturbed or a very sensitive technique used. If the detection technique is too sensitive, then hydrocarbons from a material such as rubber backed carpet could be detected and wrongly interpreted as indicating the presence of an accelerant.

Most of the volatile hydrocarbons found in the headspace of the common accelerants are also found in the headspace above most burnt plastics and synthetics but accelerant hydrocarbons are found in different ratios. A chromatographic fingerprint prepared in the laboratory must be used to determine if the hydrocarbons came from an accelerant.

The extraction technique used in the laboratory to prepare the sample for chromatographic analysis also relies on sampling and concentrating the headspace above the debris. During the extraction process it is important to recover as much of the heavier components of the accelerant as possible to avoid analytical discrimination. Extraction techniques such as purge and trap, which rely on a modified version of steam distillation give the least amount of discrimination.

Methylated spirits and acetone are also used as accelerants however they differ from the common accelerants as they are water soluble and composed of essentially a single compound. Being water soluble, they are frequently washed from the fire scene by the fire fighting operations. They are also common pyrolysis products so their presence in debris must be quantitatively assessed,

The techniques used for detecting accelerants are outlined below and discussed. The techniques are listed according to the author's opinion of their degree of usage and relative merit.

1. Physical Indicators

Physical indicators are listed first as they prove the accelerant was present at the time of the fire and was not placed there after the fire was extinguished. Investigators armed with even the most sophisticated hydrocarbon detector should not overlook the physical evidence.

Physical indicators used to detect the presence of accelerants are localised burn patterns to floors and surfaces and overhead damage inconsistent with the naturally available fuel. Reports from fire fighters or eyewitnesses of a rapid fire or of suspicious odours can also indicate the presence of an accelerant.

These physical indicators if initially present , can often be destroyed during the course of the fire. If the roof or ceiling has collapsed, then evidence such as localised burn patterns on the floor can be concealed. The investigator should excavate the debris around doorways or in the centre of open spaces as these are areas where accelerants are normally used. If a wooden floor is involved, the investigator can hit the floor with a shovel and excavate the areas where the floor appears weakened.

Physical evidence which indicates a hot and intense fire such as a colour change or spalling in concrete, meIted aluminium and deformation of steel are unreliable indicators of the presence of an accelerant, as the temperature reached during the course of a fire is governed by the amount of both fuel and oxygen available. Many combustible materials tend to burn with the same intensity as accelerants, given an appropriate supply of oxygen.

2. Use of one's sense of smell

The human sense of smell can quite correctly identify the presence of accelerants, even in trace amounts. This ability varies amongst investigators as the sense of smell is like most other senses and can become highly developed through experience, or it can become impaired both temporarily or permanently.

When one smells fire debris, they are actually sampling the headspace above the debris and noting the chemical fingerprint of the headspace. Then using one's discriminatory powers by comparing the fingerprint with those stored in one's memory, a decision can be made as to the possible presence of an accelerant.

Wine tasters use a similar technique, and their highly developed sense of smell can detect extremely minute variations in the chemical fingerprint of a wine amongst a background of water and ethanol. The same test performed by scientific analysis and scientific interpretation requires a considerable amount of time and expertise.

The human sense of smell suffers from fatigue which causes a loss of sensitivity and also the ability to discriminate accelerant vapours. The vapours found at fire scenes may be harmful so debris should only be smelt when necessary. Continually smelling these toxic vapours will cause the smelling senses to become less effective. On cold still mornings when the sense of smell is quite sharp, accelerant odours can sometimes be smelt while the investigator is making his initial inspection of the fire scene. As the debris becomes disturbed during the course of the investigation the sense of smell becomes less effective due to the contamination of the atmosphere.

If dangerous chemicals are known to be at the fire scene it is imperative to avoid smelling debris. Residues from copper chrome arsenic treated logs if inhaled could cause serious health problems. All fire scenes have noxious gases and soot particles present which can lodge in the respiratory system and cause problems. Asbestos fibres and mineral fibres from insulation are also a problem. Because of these dangers, fire investigators should not examine a fire scene soon after the fire is extinguished. At this time the concentration of toxic vapours will be at a maximum and these vapours if trapped in pockets under debris could be released during the excavation. The investigation should ideally be made a day after the fire as by that time the scene will have cooled and the toxic vapours held in pockets will have dispersed.

Cartridge respirators should be worn whenever possible during fire scene excavations, These have been considered by some to be expensive and uncomfortable but designs have improved in recent years and a good respirator can now be purchased for approximately $25 from most hardware stores.

3. Sniffers

Sniffers (or portable gas detectors) are best employed when toxic dust or vapours are present or if the investigator's sense of smell is impaired, They do not have the same discriminatory powers as the sense of smell as they respond to a wide variety of compounds in the headspace including non accelerant vapours.

A range of portable gas detectors are available on the market as industry has a need for these types of instruments to detect gas leaks or flammable or toxic atmospheres. Various types of detection techniques are employed in sniffers and the price reflects the type of detector used in the instrument.

The cheapest type of sniffer uses a detector which measures changes in the oxygen concentration. These instruments lack specificity as they respond to all types of hydrocarbons and also gases such as ammonia, alcohols, carbon monoxide, carbon dioxide and even water vapour, The advantages of using these instruments are they are small, cheap and robust. The best instruments are those which have a control to vary the sensitivity of the instrument. They can be used very effectively if the operator is familiar with the instrument and aware of their shortcomings.

A more expensive sniffer employs a detector such as a flame ionisation (FID) or photo ionisation (PID) which will respond to hydrocarbons but not inorganic vapours, The instrument is extremely sensitive but cannot discriminate between hydrocarbons originating from accelerants or those from burnt plastics. Because of their high sensitivity the investigator could easily misinterpret the results and could for example believe he is following an accelerant trail when in fact the investigator is simply following a trail where a synthetic carpet has become more severely burnt.

Sniffers do not respond to the quantity of accelerant present in the debris but to the quantity of accelerant present in the headspace above the debris. Therefore the debris needs to be disturbed before the sniffer probe is inserted amongst the debris. A large area of the fire scene can be scanned in a relatively short time by using techniques such as continuously lifting the debris with a shovel and inserting the sniffer probe under the shovel blade and noting the detector's response. Sniffers are invaluable when tracing the source of a gas leak.

4. Sniffer dogs

Sniffer dogs are used for the detection of drugs, explosives, corpse location, termites, contraband food and for tracking purposes. Dogs have a sense of smell which is much more sensitive than the human sense of smell. They also have much greater discriminatory powers and can therefore respond much more quickly to target scents. Their physical abilities and their desire to please their handlers enable dogs to thoroughly scan a large area at a fire scene in a relatively short time.

Dogs sample the headspace above the fire debris with their smelling senses and use their discriminatory powers to determine if the detected hydrocarbons originated from an accelerant . Their discriminatory powers are learnt through training.

Training generally involves a series of exercises where the dog routinely retrieves a hidden toy or object which carries the target scent. Upon successfully retrieving the object, the dog is rewarded with affection or a favourite food. When the dog locates the target, the handler notes a change in the dog's behaviour and then calls the dog so the area is not disturbed.

Dogs need to be able to discriminate between accelerant vapours and vapours such as those originating from burnt plastics and paints. They must reliably perform this task amongst a background composed of thousands of different chemicals originating from burnt furniture and building materials. Dogs should therefore be trained and rated at fire scenes.

The effectiveness of sniffer dogs is entirely dependent on the level of training the dog has been given. Drug detection dogs for instance are trained to detect drugs when a masking agent such as curry powder or pepper has been used. Criminals in an attempt to avoid pursuit by tracking dogs have been known to place urine obtained from a bitch on heat across their escape route.

Sniffer dogs have only recently been used at fire scenes but I'm sure that given the correct training they will be the greatest advancement made in recent years in accelerant detection. The use of sniffer dogs at fire scenes where the damage is widespread, such as a furniture factory fire, would be extremely cost effective as compared to using a forensic expert to dig out and inspect the entire scene.

5. Portable Gas Chromatographs

At a recent demonstration a portable gas chromatograph was used for accelerant analysis. The instrument was equipped with an FID detector and a small packed separation column. The instrument was versatile as it could be used as a sniffer where the air sample is introduced directly into the detector or as a chromatograph where the sample is introduced into the column for separation before reaching the detector and a chromatogram produced of the analysis.

The instrument was quite sensitive however the analytical column available at the time of testing was lacking in resolving power. Further developments in this area are being undertaken by the manufacturer. These developments could lead to the instrument being quite valuable to investigators as the instrument is capable of not only detecting hydrocarbons at trace levels but can also discriminate whether these hydrocarbons originated from an accelerant or from a burnt plastic. The result obtained from the machine should be verified by laboratory analysis as the machine samples the headspace above the debris which can lead to discrimination.

6. Chemical Tests

Two types of hydrocarbon chemical tests have been used for accelerant detection. Draegar tubes are routinely used for detecting hydrocarbons in the atmosphere and hydrocarbon field test kits are used for soil and water analysis and both have been used at fire scenes.

Both tests rely on a colour change as a result of the hydrocarbon reacting with a developing agent. They are generally used for the quantitative analysis of hydrocarbons and cannot discriminate between hydrocarbons originating from accelerants or those originating from burnt plastics. Both techniques are expensive to use and can only be used for a single analysis.


Of the techniques discussed above, the sniffer and the chemical tests cannot discriminate between hydrocarbons emanating from accelerants or those emanating from materials such as burnt plastics. The human sense of smell, the sniffer dog and the portable gas chromatograph have this discriminating ability.

Irregardless of the technique used to detect the presence of an accelerant the findings must be verified by the available physical evidence and by laboratory tests.