Arson Investigators: THE NEW BREED

Arson Investigators: THE NEW BREED

New York state has a new weapon in its war against arson: the Canine Accelerant Detection Program. Two Labrador retrievers trained to detect minute quantities of flammable and combustible liquids used by arsonists are part of the Department of State’s Office of Fire Prevention and Control’s (OFPC) team of experts who assist local investigators in determining the cause and origin of fires.

The program was authorized by New York’s Secretary of State Gail S. Shaffer and State Fire Administrator Francis A. McGarry following a series of meetings between the OFPC and the Atlantic City, New Jersey Police Department (ACPD). As fire investigators in OFPC’s Arson Bureau, we were assigned to work with the canines.

Under the direction of Chief Joseph Pasquale, the ACPD organized the first formal accelerant-detection canine training program in the nation. The program was developed and conducted by Sergeant Dennis McSweeney, director of the ACPD’s Canine Division. McSweeney is a 22-year veteran of the police department, including 19 years in the Canine Division training canines in patrol, narcotics detection, and explosives detection. Officer Joseph Schafer, a 13-year veteran of the ACPD’s Canine Division, also assisted with the program.


Prior to the program, Sergeant McSweeney and Officer Schafer screened approximately 500 canine candidates. Five were chosen to participate based on their play drive (motivation), temperament, ability to interact with people, and olfactory sense (sense of smell).

Training was based on the principle of praise reward—the dogs would work for the praise and love of a human being. Therefore, we each were paired with a canine so that each investigator/canine team could establish a bond of mutual trust.

The investigator/canine teams of Richard Rogozinski and Hershey and Michael Knowlton and Buddy.

(Photo by Office of Fire Prevention and Control.)

The 10-week training program consisted of classroom and field work in scent recognition, building and perimeter searches, environmental conditioning, clothing discrimination, physical agility, and canine care.

Scent recognition. The dogs learned to recognize the odors of accelerants, including seven flammable and combustible liquids in four categories: straight chain hydrocarbons such as gasoline and kerosene, oleoresins such as turpentine, polar solvents such as acetone, and aromatic hydrocarbons such as alcohol. The dogs were taught to detect an odor by playing fetch and tug-of-war with a rolled-up towel containing a minute quantity of a particular accelerant. The handler praised the dog each time it successfully fetched the towel and then played tug-of-war. All the while the dog had the towel in its mouth and was smelling the accelerant, it was associating the accelerant odor with play and affection from the handler.

Once the dogs learned an accelerant odor, they were taught to search for it and indicate when they found it in both raw and burned states. The dogs were taught to give a passive indication—to sit—when they detected such an odor. Sitting served two functions: It protected the dog from inhaling or ingesting any toxic material and ensured that the dog would not disturb an evidence collection site.



Throughout their training the canines also were introduced to large quantities of the same substance— anywhere from a half-quart to several gallons of a product was poured in various search locations. As with all other exercises, the canines indicated the presence of accelerant vapors and traced them to the area of origin without difficulty.

Searches. ‘Hie canine teams trained in various structures—single-family homes, factories, and warehouses, to name a few. Approximately 50 percent of the building searches were conducted in buildings that had suffered recent fire damage. This exposed the canines to actual working conditions and the different odors of a fire scene. Also, during building searches several distractions were placed in and around the search areas. Distractions included clean ash, burned styrofoam cups, burned foam rubber carpet backing, and burned asphalt siding. Although the canines showed some interest in these distractions, they did not give false indications. Canines also were exposed to other common distractions such as food odors, loud noises, animal and human feces and urine, broken glass, and dust and dirt.

For perimeter searches, the canines were taught point-to-point searches and quartering exercises. These searches were used to build the canines’ drive as they were conducted on several types of surfaces, including gravel, sand, blacktop, concrete, and dirt. The canines also learned to initiate their own search patterns.

A narrative report is completed each time a canine team search is conducted whether during training or an actual case. Such information as location of the call, date and time of the search, type of search conducted, and weather conditions at the time of search is included. Also noted are any locations where the canine gives a positive indication for an accelerant. The report also describes the search and includes any special conditions and/or distractions present. These reports must be completed after each search for possible use as evidence in court.

Environmental conditioning. The canines were exposed to many situations that they might encounter either en route to the investigation or while on the scene, such as elevators, escalators, and floating docks at a marina; a boat ride and boat-to-boat transfer approximately one mile offshore; commercial fishing boats; fire apparatus such as an aerial bucket truck, fire escapes, and suspension bridges; and a helicopter lift. The canines learned to trust their human partners, and as long as they were reassured they would do whatever asked.

Clothing discrimination. For the purpose of detecting accelerants that may have spilled on an arsonist’s clothing, the canines were taught clothing discrimination. Six to 10 articles of clothing were laid out approximately five feet from each other. One of the articles was contaminated with 20 microliters (about one drop) of liquid accelerant. The canines were then brought to the search area and introduced to the clothing. Each dog indicated near the article of clothing tainted with an accelerant with 100 percent accuracy.

Physical agility. Several times a week the canine teams were put through a strenuous agility course designed to build their strength and endurance and increase their search time. The course called for the canines to walk through a 15-foot section of pipe approximately two feet in diameter, climb a ladder, and perform several jumping exercises such as jump through a four-foot windowsill. It also required the canines to jump as high as they could up an eight-foot slated wall that was pitched approximately 75 degrees, then climb the rest of the wall to the top. Once at the top, the canines would have to reach the ground in two jumps—one approximately three feet to a platform, the other approximately five feet to the ground.



Canine care. Basic canine care and grooming and first aid, including canine CPR and veterinary techniques, also were taught. We considered every situation that canines could encounter during an investigation. Instruction covered obvious wounds, such as abrasions and lacerations, as well as the toxic substances that might be introduced to and/or consumed by a canine during an investigation. The canines were weighed daily at the ACPI) kennels. Each canine also was checked for potential medical problems at the program’s inception.


At the Office of Fire Prevention and Control’s Academy of Fire Science in Montour Falls, New York, the canines worked on an actual fire investigation and were featured in a 10-minute videotape covering perimeter, building, and crowd searches. This videotape then was incorporated with the other areas of training into a comprehensive 32-minute demonstration video of the canine program. It is available to fire investigators and law enforcement agencies throughout the country on request

Most of the canine training was conducted in extremely high temperatures and high humidity. However, at the Fire Academy the canines had the opportunity to work with frozen substances in a winter-like atmosphere.

In one exercise a two-by-four was burned with approximately 16 ounces of turpentine. It was allowed to burn freely and extinguish itself. Next it was placed in a pan approximately six inches deep, which was then filled with water. Following this the pan was placed in the freezer.

Then two other two-by-fours were burned using wooden pallets as the fuel source. The clean burned ash also was placed in a deep pan and frozen. The three boards were kept frozen for approximately 16 hours.

The next morning the three samples were brought to the Academy’s reburnable building. The boards were placed in two downstairs rooms. Each canine accelerant detection team indicated on the contaminated two-byfour but showed little or no interest in the clean ones.


The canines were called in to assist with an actual fire investigation of a single-family home located in South Corning, New York. Although arson was not suspected as the fire cause, a small burned can of lighter fluid was found near the area of origin. Four accelerant detection canines were brought into the building to conduct an investigation. All indicated in the area of the can, which was in the area of origin.

To test the accuracy of the canine accelerant detection team, a photoionization-type hydrocarbon detector was brought to the fire scene. The reading obtained from the can of lighter fluid was considered insignificant, meaning there would have been no further investigation of the area. This test proved that the canines’ olfactory senses were far superior to the equipment currently used by investigators.

Investigator Rogozinski and Hershey conduct an off-lead walk-through search.

Photo by Michael E. Knowlton.



In New York a twit-tier system is being implemented for the accelerant detection teams. Samples are taken not only when the canine gives a positive indication of an accelerant being present but also in the area of origin when the canine does not indicate. This supports either the investigator’s observation of the fire being accidental in nature or the theory that the perpetrator used nothing more than combustible materials germane to the structure to carry out his felonious crime.

While the formal training program has been completed, the canine accelerant detection teams still must train three times a week (in approximately one-hour sessions) to maintain their skill level. The canines have been introduced to 10 new substances that are in the four basic families of hydrocarbon materials. In addition to the weekly in-service training, the canine accelerant detection teams periodically must be recertified to ensure a high level of skill is maintained and to provide additional credibility for courtroom presentations.


Of the five canines in the original training program, two—Buddy and Hershey—now work with the State of New York Department of State Office of Fire Prevention and Control; two work with the Atlantic City Police Department; and one works with the Atlantic City Fire Department.

The statistics that follow are based on 88 fire investigations in 36 counties to which Buddy and Hershey responded throughout New York state in their first 18 months at work. These investigations have ranged from a series of residential fires in the city of Rensselaer, Rensselaer County, to a multimillion-dollar commercial structure in the suburban area of Buffalo, Erie County.

Of these investigations, 68 have been listed as incendiary, with 26 (39 percent) resulting in either arrest or indictment. The typical statewide arrest rate for incendiary fires without canine assistance is approximately 16 percent.

The incendiary cases in which Buddy and Hershey have assisted represent 13 fatalities and more than $76 million in losses.

One case involved a fire in two structures in Schenectady, New York that resulted in more than $ 1 million in damages. The fire, which occurred in December, was fought in sub-zero temperatures for several hours, resulting in a frozen fire scene. The building of origin was a three-story building of ordinary type construction that was in the process of being renovated into apartments. The second floor of the building suffered extremely heavy fire damage to structural members and contents and the third floor was totally consumed.

Hershey searched the building of origin and, within minutes, gave several positive accelerant indications on the second floor in the midsection of the building. Investigators focused their interior investigation on this area and proceeded to dig through more than four feet of frozen debris to reach the floor area. At the floor level they observed combustible/flammable liquid burn patterns. Hershey searched this area again and indicated in the same area. Investigators took several samples.

Hershey gives a passive (sit) indication of an accelerant on a couch.



A short time later, a suspect—an employee of the renovation company—was identified and subsequently arrested. The suspect told investigators how he broke into the building and poured a kerosene/gasoline mixture in the second floor hallway area in the midsection of the building—the exact location where Hershey gave a positive indication. The perpetrator was convicted of arson and is now serving a fourto 12-year sentence.

In this case alone, the canine team was estimated to have saved more than 100 man-hours of digging through the frozen scene, which not only hastened the criminal investigation but saved the city thousands of dollars in overtime.

The opposite weather conditions were present at a fire scene in Cayuga County in upstate New York: Arson was suspected in a mobile home fire where the owner was seen fleeing the scene moments before the home was fully involved.

The fire occured in early August when weather conditions were hazy, hot, and humid and temperatures were reaching nearly 100 degrees. Other distractions on the scene included holes in the floor where a wooden plank was placed so a canine could search the rear portion of the home; food on the stove and countertops; and garbage on the floor, which had already deteriorated and was covered with flies and maggots. Buddy was brought in to search the home but showed little or no interest in the aforementioned distractions. Buddy gave a positive indication in the living room area and samples were taken and sent to the New York State Police laboratory in Albany for analysis. The laboratory identified the sample as being positive for #2 fuel oil, also known as home heating fuel.

Our canine accelerant detection teams, still on the threshold of a brand new investigative field, already have proven their worth in the fight against arson. They have established themselves as experts in the field. It has been conservatively estimated that New York’s Accelerant Detection Canine Program has saved thousands of man-hours on fire scenes.

The services of the canines are available to public investigators through OFPC’s Fire Investigation Technical Assistance Program. Assistance is provided on request for fires that result in a fatality, exceed $200,000 in property loss, or have unsual circumstances—for example, explosions and serial fires.

For further information on the Canine Accelerant Detection Program, contact Arson Bureau, New York State Department of State Office of Fire Prevention and Control, 162 Washington Avenue, Albany, New York 12231, or call (518) 474-6746.


No posts to display