FOR IMMEDIATE RELEASE
Friday, June 19, 2009
Contacts:
Dave Wellman, Director of Communications, 304-696-7153 (wellmand@marshall.edu)
Mary Thomasson, Marshall University Forensic Science Center, 304-690-4363 ext. 207
 

Marshall University forensic chemistry professor
conducts study on kerosene residues in fire debris

HUNTINGTON, W.Va. – Marshall University received a $198,000 grant from the National Institute of Justice for a two-year project in which an associate professor is studying kerosene and medium petroleum distillate residues to build a database and develop methods which may aid in future investigations of suspicious fires.

Dr. J. Graham Rankin, an associate professor of Forensic Science at Marshall University, is conducting the study. The project, which began in January 2009, is titled “Development and Validation of a Method for Individualization of Middle Distillates and Kerosene Residues in Fire Debris.”

Kerosene and medium petroleum distillates, commonly sold as charcoal lighters, paint thinners and industrial solvents, are the second most common accelerants used in arson cases, Rankin said. Samples of these flammable liquids will be analyzed to produce a “chemical fingerprint” which will identify the individual liquids by its unique composition. He also is conducting similar research with gasoline, the most common accelerant found in arson cases.

Rankin said forensic application of the study includes being able to show the ignitable liquid residues found at the fire scene match a sample in the suspect’s possession or as residue on his clothing. “It could also be used to exclude the suspect residues,” he said. “These products are commonly used for a variety of legitimate uses so possession is not a crime. Excluding an innocent person is as important as, if not more important than, convicting the guilty party.”

The study will develop two tools that may assist fire investigations: a method to enable analysis of samples to establish a match and proving it works with real samples, and development of a database that will enable determination of statistical probability that the samples are identical.

Rankin said development of the database is important because there is currently no standard in place to assist fire debris analysts with statistical assessment of evidence, and courts are requiring forensic scientists to provide more proof in support of their expert opinions. This new standard is the outgrowth of the use of statistical matching with DNA evidence which is standardized, he explained.

Previously, courts relied on a forensic analyst’s experience to determine if the liquids were alike, but now courts expect the analyst to demonstrate an error rate and the probability of a random match, he said.

“The methodology does not exclude the possibility the suspect and real arsonist purchased the same kerosene from the same station, for example, during the same timeframe,” he said. “The analysis would be just a small piece of evidence that’s part of a bigger picture in an investigation – just like DNA and fingerprints may show an individual was at the scene but not necessarily committed the crime.”

A component of the research is to determine how many points of similarity are needed to determine if they came from a common source and the statistical probability of a match to similar kerosene that appears to match.

He said an analogy would be the millions of fingerprints in the population. “For years it was accepted that no two individuals’ fingerprints are alike,” Rankin said. “However, now there are studies underway which propose to look at these fingerprints to determine how many points of similarity are needed to prove two fingerprints match.”

In ignitable liquid analysis, no such database of analyses is available for research by fire debris analysts, he added.

Additionally, Rankin is developing a collection of thousands of samples of gasoline, kerosene, and medium petroleum distillates to analyze.  The samples are obtained yearly from across the United States to supply a variety of samples to make determination of statistical probability. His goal is to build a collection of more than 2,000 different gasolines and 1,000 kerosene and other similar petroleum distillates from across the country.

The ongoing gasoline project, also funded by the NIJ, has acquired nearly 1,000 gasoline samples so far with analysis by high resolution gas chromatography-mass spectrometry (GCMS) in progress. 

Gasoline is blended differently by refineries in different parts of the country based on seasonal temperature, elevation and local environmental regulations.  Additionally, chemical differences in the crude oil used to make various petroleum products including gasoline and kerosene affect the composition of the final product at the gasoline pump or in the can. The GCMS is used to separate and identify these different compounds.  Use of advanced statistical analysis then allows the determination of the closeness of agreement between any two samples as compared to all other samples in the collection.  This would enable the forensic chemist to state a probability of a successful match. 

MUFSC is collaborating with Dr. Peter Harrington of the Ohio University Chemistry Department in Athens, Ohio on the application of advanced statistical analysis of data.

The activities for the “Development and Validation of a Method for Individualization of Middle Distillates and Kerosene Residues in Fire Debris” program are funded by project number 2008-DN-BX-K146 through the National Institute of Justice, Department of Justice Office of Justice Programs.  The gasoline study is supported by NIJ project number 2005-MU-BX-K020.

###