Author(s): Scott Sutherland. Published on August 30, 2022.

Work in Progress

On the 30th anniversary of NFPA 921, the ground breaking guide for fire investigations, an influential new report presents a sobering assessment of just how far the discipline of fire investigation has yet to go


Randy Watson likes to tell the story of an early response to NFPA 921, Guide for Fire and Explosion Investigations. It was 1992, and the first edition of the guide had just been released. Watson had joined the technical committee the year before.

“My boss, a fire investigator, comes into my office with a copy of the document, throws it on my desk, and says, ‘You guys are going to have us wearing lab coats and pocket protectors before this is over with,’” he says with a grin. “It was because the guide was based in science. That freaked out a lot of people.”

Prior to NFPA 921, fire investigation, depending on who you talked to, tended to obscure the science in a haze of art, intuition, and folklore, what Watson refers to as the “ipse dixit” factor—unproven dogma handed down as fact, unchallenged “because it is the way I say it is,” he says. This tendency permeated much of the profession, including the teaching ranks.

“When I went to the National Fire Academy in 1980, they taught us that things like crazed glass, big shiny char, and burn-throughs to the floor, along with black smoke and bright orange flames, were all indicators of an intentionally set fire with the use of an ignitible liquid,” says Watson, a career fire investigator and director of technical training for SEA Limited, as well as a former technical committee chair for NFPA 921. “The science and research have since demonstrated that none of those, in and of themselves, mean anything related to an ignitible liquid being involved. They were anecdotal, and it’s impossible to trace how those assumptions were arrived at in the first place. My guess is that, at some point, someone got positive samples for gasoline at a fire scene and happened to see crazed glass, so they added one and one and came up with four. And that’s what was handed down.”

Thirty years after the arrival of NFPA 921 and the formal introduction of science into the codes and standards pertaining to the practice, the field of fire investigation has matured significantly in its sophistication and science-based rigor, thanks in part to a growing suite of related standards. A companion document, NFPA 1033, Standard for Professional Qualifications for Fire Investigator, first published in 1987, provides detail on the training and knowledge necessary for investigators. A document currently being created, NFPA 1321, Standard for Fire Investigation Units, will cover the minimum requirements for the establishment, structure, operation, and management of fire investigation units (see “How Are We Doing?” below). The new standard is scheduled for publication in 2024.

RELATED: Read Recommendations from the OSAC report on advancing fire investigation in the US

Despite those important steps—as well as the addition of a growing body of applied research and the emergence of technological tools unimaginable three decades ago—fire investigation is still regarded by some in the fire community as lagging. That view regards the field as a weak link in need of new requirements, additional rigor, more specialized science, and an array of essential resources, including funding for further research. A report published last year by the Organization of Scientific Area Committees for Forensic Science, or OSAC, acknowledged the progress made by the field of fire investigation while making it clear how far it has yet to go. “The field of fire and explosion investigation can now be said to be in its adolescence,” the report’s authors wrote, “but much work is required to bring it to maturity.”

Watson, one of nine authors of the OSAC report, doesn’t entirely agree with the use of “adolescence” as a descriptor for his profession. “I think that’s maybe a little unfair,” he says. “But I think we stayed in our infancy longer than we had to because of all that anecdotal information going around.” He describes the first book he was able to find on fire investigations, published in 1946. Among the volume’s recommendations was that fire investigators carry breadcrumbs that could be dipped into suspicious-looking spots at a fire scene, then chewed by the investigator to detect the presence of an accelerant. “We’ve come a long way since then,” he says. “Do we have a long way to go? Of course. But I also believe the fire investigation industry is on the right track by realizing we need more science and research to better support our work.” 

Creating NFPA 1321, a new accrediting standard for fire investigation units

When the Organization of Scientific Area Committees for Forensic Science (OSAC) considered NFPA’s documents on fire investigation, it saw a possible gap. There was NFPA 921, Guide for Fire and Explosion Investigations, which addressed how to conduct investigations, and there was NFPA 1033, Standard for Professional Qualifications for Fire Investigator, which addressed what investigators needed to know to do the job. What was missing was a document that outlined how the fire investigation system or agency should work—an accreditation standard.

A principle task of the 22 OSAC subcommittees is to identify gaps in discipline-specific forensic science standards. The OSAC subcommittee on fire and explosion investigation proposed a new standard to NFPA, and work was undertaken to begin writing NFPA 1321, Standard for Fire Investigation Units, currently in progress. The new standard will cover the minimum requirements related to the establishment, structure, operation, and management of fire investigation units and is scheduled for publication in 2024.

“As an accrediting standard, it will evaluate how the work is done,” says Phil Crombie, Jr., who manages the forensic laboratory for Travelers Insurance and is a member of the OSAC fire investigation subcommittee. “Most of the labs that fire investigators send their fire debris samples to for evaluation for ignitible liquids are accredited, with the process looking primarily at the quality of the work and how that quality is managed. That’s the hope of NFPA 1321 for fire investigation units.”

That agency-level rigor is essential if fire investigation units hope to make a case for the resources they deserve, which has traditionally been a challenge, says Randy Watson, director of technical training for SEA Limited and chair of the NFPA 1321 technical committee. “The standard is designed to help push the agencies to provide the necessary support for equipment, training, safety, and resources for fire investigation,” Watson says. “We need something to put pressure on the decision makers to do their jobs to make sure the agency is supported.”

For more information on the standard, visit —S.S.

The Emergence of NFPA 921
On February 17, 2004, Cameron Todd Willingham was executed by lethal injection at the Texas State Penitentiary in Huntsville, Texas. In 1992, Willingham had been tried and convicted on three counts of capital murder in the arson deaths of his three young children, who died in a fire at the family’s home two days before Christmas in 1991. Willingham insisted on his innocence during the trial—he turned down the offer of a life sentence in exchange for a guilty plea—and throughout the appeals process. He was 36 when he was put to death.

Shortly before Willingham’s execution, Gerald Hurst, a noted chemist and fire investigator, reviewed the evidence gathered by the state and found no credible grounds to support a finding of arson. “The whole case was based on the purest form of junk science,” Hurst said at the time. “There was no item of evidence that indicated arson.” Subsequent reviews supported Hurst’s findings. In 2009, the state ordered a re-evaluation of the Willingham case, and the Texas Forensic Science Commission asked Dr. Craig Beyler, then technical director at Hughes Associates (now Jensen Hughes), to conduct his own review of the evidence. Beyler’s assessment was even more damning than Hurst’s. “A finding of arson could not be sustained,” Beyler determined. His report described key testimony from a fire marshal as “hardly consistent with a scientific mindset and … more characteristic of mystics or psychics.” The commission issued a report in 2011 recommending additional education and training for fire investigators, as well as procedures for reviewing prior cases of suspected arson.

The ripple effects of the Willingham case didn’t stop there. In 2013, shortly after taking over as the Texas state fire marshal, Chris Connealy connected his office with the Innocence Project of Texas to review more than 1,000 arson convictions, a controversial effort that included the creation of a scientific advisory panel made up of some of the country’s most respected arson experts. “What’s happening here isn’t just in Texas, it’s all across the United States,” Connealy said at the time. “We need to make sure what we’re doing is indeed supported by up-to-date science.” While the arson debate was just one aspect of the tangled and problematic trial that led to the conviction and execution of Cameron Todd Willingham, there is a widespread belief throughout the fire science community that he was wrongfully put to death for the murder of his children, and that the evidence used in that conviction was the result of flawed, outmoded, and discredited approaches to conducting fire investigations.

The house in Corsicana, Texas, that was the site of a deadly 1991 fire that led to the execution of Cameron Todd Willingham for murder in 2004. Fire investigation experts have discredited the state's finding that Willingham intentionally set the fire that killed his three young children. GETTY

It is somehow appropriate that the early evolution of NFPA 921 should so closely parallel the long, high-profile drama of the Willingham case—precisely the kind of circumstance the guide, through the application of science and standardization, is intended to prevent. When the first edition of the guide was released in 1992, it spanned 119 pages over 14 chapters. It also took seven years to produce, in part due to a protracted tug-of-war between science advocates and stakeholders who believed the old, established ways of conducting investigations were not just sufficient, but preferable. “There was a lot of controversy,” says Phil Crombie, Jr., who manages the forensic laboratory for Travelers Insurance and is a member of the OSAC subcommittee. “But all of that work really advanced the state of fire investigation.”

The struggle over science wasn’t finished, however—early revision cycles were marked by attacks from the old guard that viewed NFPA 921 as a threat rather than an advancement. “For one of those editions, we had something like 150 public inputs to take science out of NFPA 921,” Watson recalls. “The inputs were identical—same wording, same grammatical errors—and most came from the public sector. We’d always talked about science, like the temperatures things burn or melt at, but now that we’re saying ‘use the scientific method’ in the guide, everybody freaks out. But we got that right—the scientific method is the right one to apply.”

The current edition, released last year, contains 29 chapters and 450 pages. More than 500 public inputs were submitted. The guide also includes a wealth of color photographs—the only NFPA document to do so. Watson argued for color when he became chair of the technical committee in 2008, along with an insistence that any new piece of information that went into the guide be supported by science. “We wanted NFPA 921 to be more of a teaching document, and the color helped us better articulate and demonstrate what we were trying to say,” he says. “A friend of mine used to describe NFPA 921 as being like a sex-ed manual without the pictures—if you’re trying to show burn patterns in a black-and-white photo that’s been copied and scanned, it’s hard to differentiate the lines of demarcation, whereas if you have a color photograph, it jumps out at you. That completely changed the dynamic of the guide.” Color photographs have been included in NFPA 921 since the 2011 edition.

A critical component in the maturation of NFPA 921—and in the field of fire investigation—is the development of guidance on burn patterns. The patterns—V-shaped, hourglass, doughnut, and inverted cones on walls and objects, to name a few—left on surfaces and materials subjected to the effects of fire amount to a fingerprint of the event. Knowing how to analyze those patterns is an essential skill for determining the origin, movement, and cause of a fire. The role of ventilation in creating burn patterns was long misunderstood, experts say, and resulted in misinterpretations of fire scenes. The more a fire progresses, the more damage tends to occur at ventilation points such as doors and windows, making it difficult to determine where the fire actually started. Research into the impact of ventilation on burn patterns produced an “origin matrix” that is now part of NFPA 921, and ventilation remains a high-priority area for further research, according to the OSAC report.

“That was one area where we had it wrong,” Watson says, of the conventional wisdom around burn patterns and their creation. “But the research caused us to look more critically at the problem and to ask more questions. And one of those questions was, ‘What does the science say?’”

Sobering Assessments
In 2009, The National Academy of Sciences (NAS) published “Strengthening Forensic Science in the United States: A Path Forward,” an evaluation of the needs of the country’s forensic science community, including fire investigation. The report was commissioned by Congress in the wake of DNA testing that revealed hundreds of wrongful criminal convictions nationwide, many of them due at least in part to forensic testimony that was inaccurate or simply wrong. Aside from DNA, the report concluded, “no forensic method has been rigorously shown to have the capacity to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual.” It also noted that lawyers and judges are insufficiently trained in scientific methodology, while arguing that police and prosecutors have too much control over crime labs, creating conflicts of interest.

The NAS project was the precursor to Department of Justice funding, administered by the National Institute of Standards and Technology (NIST), intended to address the lack of discipline-specific forensic science standards. In 2014, NIST created OSAC and its 22 subcommittees to draft proposed standards and submit them to standards development organizations, including NFPA. The subcommittees also track and publicly share the research and development needs they encounter as part of conceptualizing those standards.

With the publication last year of “Strengthening Fire and Explosion Investigation in the United States: A Strategic Vision for Moving Forward”—the echo of the NAS report is deliberate—the OSAC subcommittee on fire and explosion investigation made it clear how much forward movement it believes the discipline requires. The 150-page document, which Crombie describes as a “consensus report,” is the result of an “in-depth examination of the past and current state of fire and explosion investigation and its relationship to the judicial system,” according to the authors, who include Craig Beyler and Chris Connealy, both of whom played roles in the Willingham saga. While characterizing the discipline as adolescent in its evolution, the report also articulates recommendations for how to improve the practice (see “Steps Ahead,” previous page).

Among the key findings is that judges and lawyers “need to be educated and trained on fire science and investigation to assure that only reliable fire investigations and analyses are presented to juries.” Education requirements for investigators need to be elevated. An accrediting standard is necessary for fire investigation units—a recommendation that has already resulted in the ongoing work to write NFPA 1321. Fire investigation units need to implement information management practices. And sustained funding for research is essential for moving the discipline forward.

Research receives significant attention in the report. “Fire investigation, like many other forensic science disciplines, suffers from an inadequate research base,” the authors wrote. “Within the fire investigation discipline, there is currently little rigorous scientific study establishing the strengths and limitations of various methods and procedures. There is also little information to identify sources of bias and the impact of variations in these processes. Existing guidance in NFPA 921 does not fully address the uncertainty, repeatability, and limitations of the current methodologies.” A 16-point list of research agenda items was created, including investigating the effects of ventilation on fire patterns related to fire origin; investigating the impact of room size on the patterns that are generated; and developing techniques to limit the exposure of fire investigators to potentially biasing information.

A number of the research items address the process of origin determination, an aspect of fire investigation “that actually has had the least amount of research put into it,” according to Crombie. “Sometimes there’s still more art than science that goes into interpreting burn patterns.”

Education also gets a hard look in the report. “It has become clear that the current NFPA 1033 educational requirement of a high school degree can no longer be accepted,” the report states. “The knowledge areas that bear on fire and explosion investigation are too sophisticated for investigators to master the needed science and technology working simply from a high school education. Fire is a collection of complex physical and chemical processes that necessitates investigation by those educated in the science and practice of fire and explosion investigation. The day when a BS in one of the physical sciences, engineering, or fire and explosion investigation is the entry-level educational requirement needs to be realized in the coming years.”

Crombie emphasizes the growing technical complexity of the discipline as an important driver of the recommended educational requirements. “Fire investigation is a job of putting puzzle pieces together, and you really have to understand how those pieces go together before you can do an appropriate job of interpreting what’s going on,” he says.

Watson, though, questions the real-world utility of a four-year degree as a way to push the profession forward. “I’m all in favor of more education, but if you make a bachelor’s a requirement, then you instantly disqualify a significant portion of the fire investigation community,” he says. “And if we’re going to require a bachelor’s, then a four-year degree in what? One of the applied sciences? Well, that disqualifies me—my bachelor’s is in criminal justice. But I’ve had thousands of hours of fire investigation training since then. We need to keep looking at how we define education.”

The education debate underscores a new tug-of-war that has emerged among those attempting to shape the profession: not science vs. received wisdom, but rather an academic-leaning science approach contrasted against real-world practitioners, the people Watson describes as “dirty-fingernailed fire investigators.” Those differences have resulted in disagreements among members of the OSAC panel, he says, though they remain unified on their foundational belief: that the profession’s status quo is not acceptable, and that better processes are essential for the discipline to mature and evolve.

That includes the fundamental process of demonstrating for decision makers that the practice of fire investigation isn’t just a worthy undertaking, but an essential one, Watson says. “All the money in the fire service goes to big red trucks and fire prevention—that’s the sexy stuff that everybody sees, and the fire investigation unit gets the scraps,” he says. Watson cites one of the largest fire departments in the country, saying it once budgeted a total of $3,000 for the annual support of its fire investigation unit. “What can you do with that? Not much. Getting that support and encouragement from inside the fire department is just as important as educating the legal system, or getting more money for research, or getting rid of the junk science. That’s a goal we can all work toward.” 

SCOTT SUTHERLAND is executive editor of NFPA Journal. Top photograph: Investigators at the scene of a structure fire. Getty Images