Case of the Month: Forensic arson detection

Fire, image AdSCase of the Month: Forensic arson detection and its role in wrongful convictions feature in April.

Cameron Todd Willingham

In 2004, Cameron Todd Willingham was executed by the state of Texas for the 1991 fire that killed his three daughters: Amber (2) and the 1-year-old twins Karmon and Kameron. The prosecution argued that the fire was arson but three reviews of the evidence by arson experts concluded that it was accidental fire. Despite that, the Texas Board of Pardons and Paroles denied a petition to vacate Willingham’s conviction in 2014. The Board’s indifference in this case is a disgrace. And now we can add another disgrace: recanted jailhouse testimony and prosecutorial misconduct.

Non-disclosed Jailhouse Testimony

Johnny E. Webb was a key prosecution witness who gave testimony in exchange for a reduced sentence. From the Texas Tribune: “In a previously undisclosed letter that the witness, Johnny E. Webb, wrote from prison in 1996, he urged the lead prosecutor in Willingham’s case to make good on what Webb described as an earlier promise to downgrade his conviction. Webb also hinted that he might make his complaint public.

Within days, the prosecutor, John H. Jackson, sought out the Navarro County judge who had handled Willingham’s case and came away with a court order that altered the record of Webb’s robbery conviction to make him immediately eligible for parole. Webb would later recant his testimony that Willingham confessed to setting his house on fire with the toddlers inside.”


Case overview

On December 23, 1991, a fire broke out in one room in the Willingham home (Texas). Willingham’s then-wife and the mother of his three children, was not at home. After noticing the fire, Willingham tried to save his children but he was forced back by the intensity of the fire. He escaped the house with minor burn injuries. The investigation that lead to his capital punishment charge and conviction was based on analysis and rules that were later discredited as outdated. Willingham always denied that he had set the fire.

Forensic arson detection

At the time of Willingham’s clemency request, forensic arson detection had made significant progress. Reviews of the case showed that many conclusions drawn in 1991 were simply wrong. The most important mistakes centered on:

  • puddle configurations
  • pour patterns
  • burn trailers and
  • V-shaped burning marks

At the Willingham trial, V-shaped burn marks were interpreted as THE sole indicator pointing us to the spot where the fire started (e.g. point of origin). Now we know that V-shaped burn marks occur repeatedly during flashover and do not exclusively tell us where the fire started.

Flashover happens when the radiant heat of the flames changes a fire in a room to a room on fire. Read it twice because there is a difference. Flashover can happen within minutes and WITHOUT the presence of a liquid accelerant.

Everything in the Willingham room ignited and flashover occurred. The blaze then went from a fuel-controlled fire (due to the ignited furniture and other items) to a ventilation-controlled fire. Allow me to explain the latter. When Willingham opened the front door to escape more oxygen was sucked into the house and made its way towards the room on fire. The fire then followed the path of that new source of oxygen leaving a burn trail to the opened front door (e.g. ventilation-controlled fire). THIS burn trail was mistakenly believed to be caused by a liquid accelerant.


After many unsuccessful appeals, Willingham’s last hope was executive clemency by then-Governor Perry. Perry received the Willingham defense report with the new scientific analysis but denied clemency. Willingham was executed for an accidental fire.

This isn’t the only case where accidental fire was mistakenly determined to be arson.

Frontline reported on March 18 that “this month, the State Bar of Texas formally accused the lead prosecutor of misconduct after an investigation requested by the Innocence Project.”

The quest for better forensic arson detection and investigation continues.

Daniel Dougherty

In 2000, Daniel Dougherty was found guilty of the 1985 arson that killed his two sons, Danny (4) and Johnny (3). Police arrested Dougherty in 1999 after his estranged wife and mother of the two children claimed he had confessed. A jury found him guilty on capital murder charges in 2000. He is awaiting his execution.

The Pennsylvania Supreme Court affirmed Dougherty’s death sentences in 2004. In 2005, Dougherty filed a petition for a writ of certiorari to the US Supreme Court. Certiorari was denied the same year. Dougherty is an inmate of the State Correctional Institution at Greene.

Case overview

On August 24, 1985, Dougherty skipped his Alcoholics Anonymous meeting and went to a bar instead. He got into an argument with his then-girlfriend, went home, made dinner, and fell asleep on the sofa in the living room. When he woke up he saw that the curtains were burning. His boys were asleep upstairs.

According to trial transcripts, Dougherty ran outside to grab the neighbor’s garden hose. However, it wasn’t long enough to reach back to the fire. Very soon flames burst from the house. Glass exploded cutting his arm. He grabbed a wooden ladder to reach the boys but he testified the fire “blew him down.”

When firefighters arrived, Dougherty was frantic and screaming at police to save his children. His aggressive and erratic behavior did not sit well with them. They shoved his face in the mud and took him away. Dougherty testified that at that moment, he just wanted to die.

Authorities who sifted through the charred remains determined that the fire was intentional. Police questioned Dougherty and his family members but no arrests were made. Dougherty’s alcohol addiction intensified as he tried to cope with loss. He eventually divorced his wife, remarried, and divorced again.

In 1999, police paid Dougherty a surprise visit. His second wife, Adrienne Sussman, had informed police that Daniel had confess to have used gasoline in the fire. Dougherty was promptly arrested. The defense argued that Sussman’s claim should have been dismissed immediately because no fire report stated that any accelerants were used to start this fire. Sussman went to police at the most opportune time. According to court documents she was fighting Dougherty for custody of their son, Stephen.

Despite the discrepancy between the fire analysis and a self-serving motive, the prosecution went ahead with the trial and boosted their case with jailhouse testimony. Two inmates testified that Dougherty had confessed to them. The defense of course objected pointing to studies that show how in-custody informant testimony is one of the leading causes for wrongful convictions in capital punishment cases.

Dougherty’s first wife and the mother of the deceased children, Kathy Fox, said that she didn’t believe that her ex intentionally killed their children. She never testified in the original trial because Dougherty’s attorney made a tactical mistake and didn’t call on her. “Knowing Daniel and his relationship with his children, I cannot believe he would have burned them to death,” she said later in statement presented at Dougherty’s appeal.

Intentional or Accidental?

John Lentini and Angelo Pisani (two of America’s best known arson investigators) separately reviewed the Dougherty case. Neither found signs of arson. In their 2005 report, they emphasized the use of outdated fire investigation techniques. Then fire Marshall Quinn thought that there were three fires: near the sofa, one near a love seat, and the third one under a dining room table. Somehow he also thought that only one person could have set those three fires. I do not know why.

FireCompare Willingham with Dougherty

In 2004, Lentini reviewed the Willingham case. Just like in the Dougherty case, he didn’t see signs of arson. He did see signs of flashover. “Flashover fires can be mistaken for arson because they leave the appearance of multiple points of ignition.

About the Dougherty case, Lentini said that Pennsylvania is “on their way to executing an innocent man. The origin of the fire could not be determined because of extensive damage to the room.”

Multiple points of ignition also plagued the Willingham case. There were three: one was in the hallway, one inside the girls’ bedroom, and one was near the front door. The 1985 the conclusion was that only a human could have done that.

Kenneth Richey

The Dougherty and Willingham cases show exactly the same issues that plagued the Kenneth Richey case. Richey was on death row in Ohio for arson with transferred intent. Richey allegedly had the intention to kill his former girlfriend, Candy Barchet and her new boyfriend, by setting fire in the apartment ABOVE theirs. Richey was sentenced to death.

Affidavit Richard L.P. Custer regarding the burn pattern in the Richey case

Affidavit Richard L.P. Custer regarding the burn pattern

The apartment above his ex-girlfriend’s belonged to Hope Collins who lived there with her daughter Cynthia (2). On June 30, 1986, Hope left Cynthia alone to go to a party. The prosecution claimed she asked Richey to babysit but instead he left Cynthia alone.

On appeal, Richard Custer, a specialist in fire reconstructions, testified in court that the burn patterns found in Hope’s apartment could have been a natural result of the fire. He also testified that to create the accelerant’s pour pattern and location that the prosecution pointed at Richey would have to use about 10 gallons of fuel. According to the prosecution Richey climbed on top off a utility shed right below Hope’ living room balcony, then he climbed over the railing, and used the accelerants to set the apartment on fire. The problem with this is that Richey did all that with a broken hand in a cast and without getting one drop of accelerant on his clothes and boots.

The prosecution admitted at trial that they could not prove that Richey intended to kill Cynthia. They also had to acknowledge that Richey, when he saw the building on fire, made several attempts to save the child but that he was held back by flames, smoke, and ultimately, by the fire department.

Affidavit Andrew T. Armstrong mentioning how the living room carpet was retrieved in the Richey case

Affidavit Andrew T. Armstrong mentioning how the living room carpet was retrieved in the Richey case

The local Fire Chief and the Ohio State Fire Marshall ruled the fire accidental. The prosecution then hired their own experts. They were sure that the fire was ignited with stolen paint thinner and gasoline. Traces were found on Hope’s carpet.

The problem here is that this part of Hope’s carpet was found at the dump. It was partly covered by trash. When it was removed it was taken to the Sheriff’s department. There is was left in the parking lot for three weeks within approx. 35 feet from their gasoline pumps before they took it to a lab for testing. The defense failed Richey here by not having their own forensic tests done.

In 1996, arson experts Andrew T. Armstrong and Richard L.P. Custer reexamined the evidence. They found that the original findings were based on outdated and unsound scientific principles. You can read about the Richey case here as there is a lot more to this case than I just mentioned. Also, please read what Tony Cafe, an Australian forensic scientist, reported about the evidence that was used to convict Richey.

Evolving science

Forensic arson detection evolved but not everyone kept up with the progress. According to some in the industry, before 1992 many guidelines analyzing arson were “largely based on hand-me-down myths” and many fire investigators lacked formal training. In 1992, the National Fire Protection Association released the first arson guidebook NFPA 921 one year after the Willingham fire

The widespread attention for DNA on TV and in the movies has helped to make it as well-known as it is today. But that didn’t happen with forensic arson detection.

Kristine Bunch

Nobody knew why an Indiana single, unemployed mother on food stamps without life insurance policy would kill her son by setting fire to the house. Bunch was convicted of arson that killed her son Tony (3). The prosecution’s case was based on the presence of accelerants, that gasoline was found close to the front door, and that there were burn patterns. Bunch received two concurrent sentences of 60 years for murder and 50 years for arson.

Bunch got help from the Center for Wrongful Convictions at Northwestern University. From Rob Warden: “Three fire forensic experts — Jamie McAllister, John DeHaan, and John Malooly — who concurred in the view that the arson testimony presented by the prosecution at Bunch’s trial in all likelihood had been wrong. One of the first things they did was subpoena ATF files on the original investigation.

In response, the ATF surrendered previously undisclosed documents showing that — contrary to the trial testimony of William Kinard, the ATF analyst — no heavy petroleum distillate had been found in the bedroom. In 2012, the Indiana Supreme Court unanimously declined to disturb the Court of Appeals decision. However, Bunch was released on her own recognizance 24 days later after she served 17 years. Shortly before Christmas 2012, the prosecution dropped the charges.”

David McLeod

There will be no retrial in the 1989 arson case that killed the Hina Family after the prosecution decided not to retry David McLeod. Right now, McLeod is free but can be retried if new evidence or legal avenues develop.

McLeod was held without bond since his arrest in 2010 (Keene, New Hampshire). A mistrial was declared in 2013. A retrial was scheduled for April 14, 2014. “Attorney General Joseph Foster said it became clear after a review, which included juror interviews, that a different outcome after a second trial was highly unlikely. Therefore, after a thorough review, the decision has been made to forego the April 2014 retrial and avoid a double jeopardy issue,” Foster said in a statement.”

McLeod was accused of setting fire to a multi-unit apartment building I Keene, New Hampshire, in January, 1989. It killed the Hina Family. The medical examiner ruled they all died of smoke inhalation.

Here too flashover occurred as in the earlier cases as well as the use of outdated science:

  • We thought that fire fueled by accelerants burned faster and hotter than other fires
  • We thought that crazed glass, V-patterns, and black markings were evidence of pour patterns from accelerants

We were wrong.

Richey’s case is similar to McLeod’s. In both those cases a burning or smoldering cigarette might have set ordinary household pieces that naturally contain combustible materials on fire. In both cases the question was whether Richey or McLeod had added any accelerants. In Richey’s case, that was never proven.

FireThe Daubert Standard

There are many cases where accidental fires have been ruled arson sending people to prison or worse, to death row. On this website, you can read about some myths that surround arson. Pay attention to the Editor’s Note at the bottom. They note that the industry failed to “adequately police itself.” It continues to discuss the Daubert Standard.

In 1993, the US Supreme Court Case ruled in Daubert v. Merrill Dow Pharmaceuticals that the trial judge must make a preliminary assessment of the expert scientific testimony. The judge must decide if the testimony is based on reasoning or a method. If it is based on a method then that method must be scientifically valid AND it must apply to the facts of THIS case.

The factors that the trial judge must weigh are:

(1) Whether the theory or technique in question can be and has been tested;

(2) Whether it has been subjected to peer review and publication;

(3) Its known or potential error rate;

(4) The existence and maintenance of standards controlling its operation; and

(5) Whether it has attracted widespread acceptance within a relevant scientific community.

The Daubert standard is the test used in the federal courts and in some state courts. In the federal courts, it replaced the less precise Frye standard which stated that a judge had to decide whether a method was generally accepted by experts in the particular field of expertise to get evidence.

In all the above cases we were dealing with accidental fires that tragically took lives of mostly children (except in the Hina case). The cold element does not so much lie in the identity of the attacker that got away as is normally the case. Rather, this time it lies in neglecting the evolving standards of science, lack integrity, and disrespect for the law that resulted in wrongful convictions.


We do not know exactly how many people were sent to prison because of outdated science let alone due to corruption. We do know that when industries fail to monitor themselves and do not require continued education from its their members, others pay the price.

We must remain highly suspicious of jailhouse testimony. If presented in court, it should be accompanied by a thorough ethics investigation of the prosecuting attorney. The Webb letter in the Willingham case shows us again that desperate self-serving measures can lead to the conviction of the innocent and that less than ethical councils still willingly receive such “evidence” to win their case.

But above all, it is time to renew our respect for procedure and disclose all evidence to the defense to prevent wrongful convictions.




News Australia

Houston Chronicle

The Marshall Project

The New Yorker

PBS Frontline

Incendiary, the movie






Richey v. Mitchell

Tony Cafe Forensic Report

Toledo Blade

Death Penalty Information Center

Forensic Science Central



National Registry of Exonerations


The Sentinel Source

Valley News

New Hampshire Dept of Justice Office of the Attorney General

Sea Coast online

Fire Rescue


In the series “Case of the Month” I highlight old cold cases or wrongful convictions. These posts are not an in-depth analysis and of course, more information can be found online and in newspaper archives.

Most of the forensic arson related cases have been discussed extensively on my blog before but as we see with the Willingham case, they need every exposure. We must grab every opportunity to raise awareness for these miscarriages of justice.

If you have any thoughts about prosecutorial misconduct, jailhouse testimony, forensic arson detection, or one of the cases discussed, I encourage you to post on your social media platforms such as Twitter, Facebook, Google Plus, etc. or to blog about it.

Thank you for reading.


  1. […] have written extensively about forensic arson detection. In this post, you can find 5 cases where accidental fire became arson and resulted in the wrongful conviction […]

  2. […] assessment on fire investigations. As you can read on my blog, we have some cases were people were wrongfully convicted for arson that later turned out to be accidental fire. The report that came out last July discusses the […]

  3. […] This case combines a lot of factors ranging from inadequate leadership to negligence to outdated science in forensic arson detection. I wish we still had the hotel carpets so they could be examined with modern technology. You can find more details about forensic arson detection here. […]

  4. […] is similar to the cases of Kenneth Rickey, Cameron Todd Willingham, Daniel Dougherty, and others. Forensic arson detection evolved but not everyone kept up with the progress. According to some in the industry, before 1992 […]

  5. […] However, as forensic arson detection progressed we now know that ordinary pieces of furniture contain flammable materials such as upholstery (check point #8). Accidental fire was ruled out as the furnace/house had new wiring and the victim did not smoke so discarded cigarettes were not an option. Note: to learn more about forensic arson detection click here. […]