This is the third of four articles focusing on two arson cases, Willingham and Willis, and urging those who may be charged with investigating similar cases in the future to be as well prepared and able to defend their findings as possible.
This article deals with the April 15, 2011, Texas Forensic Science Commission’s (TFSC) 800-page report that resulted from its review of the testimony of fire investigators Ed Cheever and John Daily in the State of Texas vs. Ernest Ray Willis case. (Willis was exonerated on Oct. 6, 2004, as opposed to Cameron Todd Willingham, who was put to death by lethal injection that same year.) These columns highlight some of the areas on which the TFSC reviewers focused, in the hope that we fire investigators can, by thoroughly educating ourselves and copiously preparing, preempt our future cases from being disparaged and can continue to learn and improve.
Trial testimony of Edward Cheever
Edward Cheever testified on July 28, 1987, that on the day he investigated the Willis fire, he had only been an arson investigator for two years and had only been certified as an arson investigator for eight months, according to the TFSC report.
The report also said that, during the trial’s voir dire, Cheever advised he had not yet been fully trained as a lead investigator and was still in training under Leroy Brown who, although listed as the lead investigator, never testified. The report said that Cheever stated, “I assisted in the investigation.” Nonetheless, he was allowed to testify and give his opinion during the trial.
The reviewers then focused on page 14 of his testimony, in which Cheever is quoted as having said he believed that the low burning of the fire was an indicator of ignitable liquid use. They said the investigators in both the Willingham and Willis cases relied on their ability to visually interpret irregular burn patterns on floors of rooms subjected to full compartment fire involvement. It is now known, the TFSC reviewers noted, that post-flashover compartment fires commonly create irregular floor patterns. They cited test results included in Shanley, J. H.’s “Report of the United States Fire Administration Program for Study of Fire Patterns,” published July 16, 1997.
Further, page 21 of Cheever’s testimony caused reviewers to question his understanding of fire dynamics. The concept of flashover compartment fires and the characteristics of rooms subject to post-flashover conditions were well-known facts in 1986, summarized in 1985 by Drysdale, D., in “An Introduction to Fire Dynamics.” Regarding page 27 of Cheever’s testimony, the reviewers noted that low burn at a doorway and a subsequent “V” pattern, occurring in adjoining areas, were only indications of the presence of a fully involved compartment fire, regardless of the presence of accelerants. Therefore, Cheever’s assumption that the “V” pattern, the low burn, and the irregularly shaped patterns on the floor, described on page 31 of his testimony, were indicative of accelerant use revealed his lack of knowledge of research and of fire dynamics.
One of the last areas on which reviewers focused was Cheever’s lack of statistical knowledge regarding the leading cause of fire deaths in the United States at the time of his testimony: cigarettes.
Review of John Dailey’s testimony
On July 29, 1987, John Dailey, a retired FBI agent who was, at the time of his fire investigation, an insurance claim fraud investigator, testified. The TFSC report noted his fire investigation experience was limited to a 90-hour arson investigation course he had taken in May 1983, which certified him as an arson investigator in the State of New Jersey.
As with Cheever, the reviewers focused on Dailey’s misinterpretation of “V” patterns and low burn and his lack of familiarity with characteristics of the effects of a post-flashover compartment fire and of fire dynamics: the fact that this type of event can cause these results without the introduction of accelerants. Reviewers also highlighted page 29 of Dailey’s testimony in reference to char damage below the doorjamb; a popular misconception at the time held that a fire in a doorway was indicative of the use of accelerants, a supposition that has since been debunked.
Reviewers looked at pages 32 and 33 of Dailey’s trial testimony, specifically regarding the condition of the couch at the fire scene. They concentrated on Dailey’s interpretation of what he believed was the cause of the annealing and sagging of the couch springs on one side: the rapid spread of the fire due to the use of accelerants. Reviewers cited research by Tobin and Monson in 1989, from “Collapsed Spring Observations in Arson Investigations: A Critical Metallurgical Evaluation,” which held that the degree of spring collapse isn’t an indicator of the speed at which fire spreads.
They also questioned Dailey’s statistical knowledge of smoking materials, the number one cause of fire deaths, which could cause similar resulting couch annealing. Reviewers again quoted Shanley, Test 6 of the USFA Fire Patterns tests, and noted that pouring ignitable liquid does increase the speed with which other fuel sources in a compartment catch fire. However, in the absence of other fuel, such as furniture, the single application of an ignitable liquid would not be sufficient to result in compartment fire flashover.
Recommendations
The TFSC report recommends continuous, remedial, advanced education and development of fire investigators. Reviewers promote fire incident investigation and fundamental knowledge in scientific physics and chemistry of fire as a prerequisite to the practical application of fire dynamics within the context of the scientific method. They believe that fire investigators must continually seek knowledge through scientific literature on fire behavior. The reviewers also believe this will improve the quality of fire investigations, allowing more individuals from the fire investigation community to contribute to available scientific community literature. These steps, reviewers believe, will lead to better quality training, educational program advancements within the profession and quality control through self-policing.
Note: In the first article, I incorrectly stated that the scientific method is a five-step process; it is actually seven steps as defined by NFPA 921, 2014.
Collin County Fire & Arson Investigators Association, which I have the honor of heading up, is sponsoring the fourth annual fire death investigation course later this year. Held in conjunction with Eastern Kentucky University, Sam Houston State University and the Texas State Fire Marshal’s Office, it is slated for Sept. 26–30, 2016, at Sam Houston State University in Huntsville, TX. To participate, download the packet at http://tinyurl.com/oeb2ys9 or contact me at [email protected].
STEVE SEDDIG has been a member of Wylie, TX, Fire Rescue since 1994 where he currently serves as fire marshal and division chief. He is the elected president of the Collin County Fire & Arson Investigators Association and was instrumental in bringing to Texas one of only two fire death investigation courses in the country that uses human case studies. He holds a Bachelor of Science in Emergency Administration from the University of North Texas in Denton, an Associate of Applied Science—Fire Science from Collin College in Plano, TX, and Associate of Applied Science—Law Enforcement Technology from Rio Salado College in Tempe, AZ.
