LAURA SULLIVAN, HOST:
From NPR News, it's WEEKENDS on ALL THINGS CONSIDERED. I'm Laura Sullivan, in for Guy Raz.
In 1990, a fire broke out in a house in Jacksonville, Florida. Two women and four children died. The husband of one of the women became the prime suspect. Everyone thought it was arson. That's when a fire investigator named John Lentini was called in.
JOHN LENTINI: So I met with the captain and I looked at his evidence, and it looked like what passed for good arson evidence in those days.
SULLIVAN: Lentini looked for V marks on the walls to see where the fire started.
LENTINI: Shiny alligator blisters on the floor were a sign of accelerant.
SULLIVAN: He calculated it would take about 20 minutes for the house to become engulfed, something experts call flashover, plenty of time for the husband to set the fire. All the while, though, the husband claimed his innocence. Then, almost on a whim, Lentini noticed an identical house two doors down that was slated to be demolished.
LENTINI: So, what else could we do but burn it down? So, we refurbished the house and then we lit it on fire. Well, after four minutes, we had flashover in that room. And, you know, I was pretty shocked. I thought, oh, my gosh, this might not really be a set fire.
SULLIVAN: And it wasn't. They dropped the charges against the husband, and it was a seminal moment for Lentini.
LENTINI: What I knew about arson, some of it was wrong. And what a lot of people thought they knew about arson was wrong.
SULLIVAN: That's our cover story today: the changing science of fire investigations and how what may once have been considered science wasn't, in fact, science at all.
LENTINI: The only science that anybody thought about was the science that happened in the laboratory when chemists would look for gasoline on the living room rug. The rest of it, it was art.
SULLIVAN: Where did all of these old ideas of how fires started come from?
LENTINI: Anecdotal experience. I mean, nobody ever set out to send an innocent person to jail. But if somebody would see an artifact and then find gasoline, they would make a connection. And then the next time they'd see that artifact, they just assume that gasoline must have caused it.
SULLIVAN: Do you think that means that innocent people have gone to prison?
LENTINI: Oh, absolutely.
DOUG STARR: The few months of research I did on this turned up at least two dozen cases.
SULLIVAN: That's Doug Starr. He's the co-director of Boston University's Center for Science and Medical Journalism. And he recently wrote an article for Discover Magazine examining arson cases that relied on now-widely debunked theories about how fires start.
STARR: One of the earliest cases was a 16-year-old named Louis Taylor. He was convicted of setting the infamous 1970 Pioneer Hotel fire in Tucson. He was convicted on 28 murder counts based on the then fire investigator's observation that at opposite ends of a hall, there were severe areas of burn.
At the time, it was thought that the area of the deepest burn must be where the fire started because it burned there the longest. And working backwards from that evidence, they figured that because they found Taylor with several matches in his pocket, he must have set those two fires and they must have been set independently.
Last year, at the behest of an attorney, several fire investigators looked back at the evidence and they said this looked like a classic accidental flashover fire, which you do get deep burn areas in separate areas not because they were set but because that's what ventilations sometimes does. But Taylor is still in jail.
STARR: Yeah. The Arizona Justice Project is trying to get the case brought up. But these cases are extremely hard to get re-reviewed.
SULLIVAN: How many cases do you think are there out there?
STARR: The investigators I speak to say that there may be hundreds out there. For example, you all know about the Willingham case in Texas in which Cameron Todd Willingham was executed for, you know, burning a house and killing his children. The Texas Forensic Science Commission, in reviewing that, found that that was not a good decision. And just at the end of October, they said that all cases in Texas have to review. We're talking about between 750 and 900 arson cases in Texas alone.
SULLIVAN: Have you noticed any common threads?
STARR: Yes. First of all, it must be said that everybody means well. Everybody's doing the best that they could do. But for generations, fire investigations have been based on kind of folklore, apprentice-type knowledge in which a fireman may be consigned to the investigative unit and he or she will learn classic signs of arson from the person who came before. The trouble is this was all based on observation and intuition. And science needs something more than that. And fortunately, the new forms of investigation that are emerging are based on actual laboratory science in which things are demonstrated to be true or not true.
SULLIVAN: What are some of the new techniques that are being used now?
STARR: It's really interesting. Scientists are analyzing smoke detectors. It turns out that when a smoke detector goes off, you know that high-pitched siren sound it makes?
STARR: That causes a kind of sound vibration that causes tiny particles to accumulate in particular patterns around the sounding horns. So by analyzing that, they could get a better idea for how long a smoke detector was going on or if it went off. So, this is all science. It's been quantified in the laboratory. People can attach numbers to it. It is outlined in professional standards. And given time and money, people can analyze a fire pretty well nowadays.
SULLIVAN: Some of that newest research on how fires start and spread is now coming from ATF. The federal agency's always done a little fire research. But it went all in in 2003 with a new lab in Beltsville, Maryland, built just to burn things up.
JAMIE LORD: He's measuring out the diesel fuel now into our little graduated cylinders here.
SULLIVAN: Jamie Lord is one of ATF's fire research engineers.
LORD: You put a little gasoline in there, Jimmy?
LORD: Yeah, he spiked it with a little gasoline. So if you poured diesel into a pan and tried to light it with just a match or a lighter, you'd be here all day.
SULLIVAN: ATF runs these experiments all day long to study heat, combustion, burn times.
SULLIVAN: That's really a lot of smoke for just a small pan.
SULLIVAN: The researchers say they're pretty used to smelling like they're at a never-ending barbecue. John Allen is the chief of fire research here and he gave us a tour.
This is a very large room. This is like an airplane hangar.
JOHN ALLEN: This is the largest forensic investigative tool in the world.
SULLIVAN: And I have to say, this is the largest upside-down cone I have ever seen in my life.
ALLEN: Well, basically, this is a 40-by-40-foot inverted cone over our head, which when we set a fire here acts to pull that smoke out of the room, and the heat, to measure carbon dioxide, carbon monoxide and things of that nature. And what you see here is actually a quarter-scale model of a room...
SULLIVAN: It's a living room.
ALLEN: It's a living room with a couch, a couple of chairs, a TV set on a small bookcase and a crib with a few stuffed toys in there. As we set a fire in this room, what will happen is the fire will begin to grow. Cool air comes in from the bottom. The hot air and hot gases will start escaping out of the top. As the fire gets bigger, more and more heat will start pooling at the ceiling and fire will start extending out, but then the heat radiating down from all that burning gas above will start catching everything in the room on fire, which is a transition phase that we call flashover, which is essentially going from a fire in a room to a room on fire.
Once you go into a flashover stage, all the original patterns and things like that on the wall can now be changed and obliterated. You can't walk in and look at the scene and say, well, I know the fire started here because of this damage. Because, no, the whole room was so severely damaged that you have to do much more processing.
SULLIVAN: Is that different from the way fire was understood 10, 20, 30 years ago? Do you think it's made sort of leaps and bounds in the past 10 years?
ALLEN: I would say because of the presence of this laboratory, yes, there has been advances in scientific knowledge. For example, if I take a cigarette and throw it in gasoline, it'll go out.
SULLIVAN: Oh, but they have it in all the movies!
(SOUNDBITE OF LAUGHTER)
ALLEN: That's TV. That's exactly right. That's the "CSI." Exactly, that's TV.
SULLIVAN: That's how you make a gas tanker blow up, right?
ALLEN: And I'm sorry I'm going to dispel that myth, but we have done thousands and thousands of tests with cigarettes and gasoline to try and get cigarette to ignite gasoline. We have not been successful yet. But it's - I didn't say it's impossible, so don't try it.
(SOUNDBITE OF LAUGHTER)
ALLEN: Because you may be the one that actually finds that, hey, I did get it to happen.
SULLIVAN: What do you think are some of the great - other great myths that you've dispelled in this lab?
ALLEN: Oh, a couple of old ones, spalling of concrete, which is...
SULLIVAN: What is spalling of concrete?
ALLEN: If you look at a concrete floor after a fire and you see chunks and slabs all broken up across the concrete, well, that was an indication that somebody poured gasoline on there.
ALLEN: No. It just means there was moisture in the concrete. And we all know now that that myth from 40 years ago is not true.
SULLIVAN: Next to the living room is another structure, an almost completely finished trailer. John Allen says this is the other half of what ATF does now, it recreates fires to help local investigators with actual cases where arson is suspected. In a few days, they're going to burn this trailer down.
ALLEN: This used to be the combination kitchen area and living room area where there was a couch. We had a mock-up refrigerator. We actually go out to wherever we can to find the materials like the couches and things like that to bring them in here and set the fire up.
Some cases, like this room was not actually involved in a fire, but because the smoke and the heat progressed to the trailer, then we have to build all those different areas to include them in there, to mimic what was really there. We have our crew assembled with all the materials we need here to quickly put things together and to answer those specific questions that may come up. Again, we're not the actual fire investigators. We're the engineers and scientists that support...
SULLIVAN: Yeah. Do you ever get tired of building things that you're just going to burn down?
ALLEN: I couldn't think of a more fun job in the world, to be able to come in and every day, we're doing something new and different. And many days, it's something cutting-edge that's never been done before.
SULLIVAN: This year alone, this lab has recreated fires from three murder cases. In all of them, prosecutors ended up dropping the charges against the suspects, because the lab determined that what officials thought might have happened when the fire started actually didn't.
NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR's programming is the audio.