July 1, 2018 Vol. 1, No. 1

The Lime Street Fire Test, 1991. Soon to be on the cover of

Scientific Protocols for Fire Investigation, Third Edition

How Judges Decide Expert Challenges

This article from JDSupra discusses the admission of expert testimony, including fire experts. It’s not a very flattering picture. There is a long discussion of NFPA 921 beginning at page 27.


An Account of the Moonlight Fire

A new book by Joel Engle, Scorched Worth (Encounter Books 2018) describes the billion-dollar litigation by the US Government and the State of California against Sierra Pacific Corporation. I participated in this litigation as a defense expert from 2011 through 2013. The litigation continues to this day. Definitely worth the time it takes to read, this book is actually two parallel books intertwined. One describes how Archie “Red” Emmerson built Sierra Pacific into one of the largest privately held landowners in the country. The parallel book describes the flawed and ultimately corrupt fire investigation and litigation pursued by the governments. Sadly, the villains were all promoted.

Australia, Canada Make Explicit an Expert’s Duty to the Court

Although the proposition that an expert should be independent has always been accepted wisdom, the requirement that an expert be independent has only recently been mandated by the Court as part of the threshold test for admissibility.

Thanks to Richard Rice of Mutual Engineering for this link:


Case Study of the Month

This column will provide my review of a case of interest to fire investigators and litigators. The first one, from Wisconsin, has a novel take on what constitutes “new evidence.” This case is reproduced from Scientific Protocols for Fire Investigation, Third Edition (in press).

State of Wisconsin v. Joseph Awei

Joseph Awe was the owner of a bar, JJ’s Pub, located in Harrisville, WI, which was destroyed by fire on September 11, 2006. The building was over 100 years old and had electrical problems. The state’s investigators found no ignition sources at place where they thought (erroneously) that the fire started, used “negative corpus” methodology and declared the fire to be incendiary. They also made an issue of a framed poster of Joe and colleagues on an Army Jeep in Kuwait at the end of Desert Storm. The image was once on the cover of Life Magazine but in a later incarnation, it was used as part of a Miller Lite promotion. It burned in the fire, but the state’s investigators said Mr. Awe had removed it prior to the fire because of its sentimental value. This was based on an outdated notion of the irreplaceable nature of photographs. Miller actually sent Mr. Awe a new poster within three weeks of the fire. There was $900 cash found in the till, presumably left there to mislead the investigators. Mr. Awe was 40 minutes away at the time of the fire, so the state proposed that he hired someone (never even tentatively identified) to set the fire. The insurance company “supported” the state’s investigator with an electrical engineer and another fire investigator. It was the electrical engineer’s elimination of electrical cause that caused the lead state investigator to write, “The conclusions section of the fire scene examination report should be amended to indicate that the official cause for this fire incident is now being classified as incendiary in nature. Electrical engineer Chris Korinek has eliminated electricity as a potential cause for this fire incident.”

Mr. Awe’s trial counsel also hired an electrical engineer, but he had never testified and he spent all of ten minutes examining the service panel. He was discredited, and Mr. Awe was convicted on December 20, 2007. He remained free on bond pending appeal until 2011, when, having exhausted all of his appeals, he began serving his three-year term of imprisonment.

The author was not contacted until the Spring of 2012, when Mr. Awe’s sentence was more than half over. When I queried the appellate attorney, Stephen Meyer, why he would expend such effort, he stated that his client wanted his reputation back.

Figure 9.4 (a) Front of JJ’s Pub

Figure 9.4 (b) View of the storage room where the fire originated, view from the bar area.

Figure 9.4 (c) Pre-fire view of the bar showing the “missing” (actually burned) poster. Blue door on the right leads to the storage room where the fire originated.

Appellate counsel also retained Electrical Engineer Mark Svare to reexamine the electrical evidence.

Figure 9.4 (a) shows the front of the bar, and Figure 9.4 (b) shows the alleged point of origin.

Figure 9.4 (c) is a “before” shot of the area shown in the foreground of Figure 9.4 (b), showing the door to the storeroom and the framed poster that the state alleged had been removed in anticipation of the fire.

Figure 9.4 (d) Base of the alleged area of origin prior to some of the wainscoting at the bottom being removed.

During the investigation, the state’s investigators tested their origin hypothesis by conducting a reconstruction. There had been a shelf mounted on the wall directly above the area where the investigators initially hypothesized the fire began. Examining the shelf, however, revealed that it was not nearly as damaged as the interior of the wall. This indicated that it fell down early in the fire, and was not exposed to a fire burning on the

floor. Rather than looking for another origin, though, the investigators modified their hypothesis to one that put the origin inside the wall. What clearly happened here is that there was fall down inside the wall, shown in figure 9.4 (d), a common feature of fires in buildings with balloon construction. The reconstruction is shown in Figure 9.4 (e).

Figure 9.4 (e) Reconstruction of a shelf above the alleged area of origin. A fire set below this shelf should have caused far more damage to the exposed bottom side of the wood.

This was the state investigator’s testimony on direct:

"You have two by four studs within the wall there that, once the fire breaks through that wainscoting or exterior surface and gets into the interior part of the wall it's gonna act just like a chimney. Those of us who have fireplaces know what happens when the fire gets up into the chimney: It's gonna go to the least area of resistance. And all that available oxygen is up high so the fire is just gonna rise rapidly in between the studs and the wall. And in balloon construction, this type of construction, older construction, it's very common. Once the fire gets into the walls it goes right up to the second floor and into the attic immediately. And that's why you see such a distinct pattern here. And then the protected areas around it are like that because there was stuff stacked up around it protecting it. This was a storage room. There was all kinds of stuff stored on shelving and on the floor around it.”ii

Now compare his testimony on direct with his testimony redirect reflecting the change in origin to the inside of the wall:

What it tells me, based on the elevation of the damage, the nature of the damage, and how it got into the wall so quickly, to me the likeliest scenario there is someone put a boot through there. Someone came up, gave it a good swift kick or knocked it out with a hammer or an axe or did something to damage it intentionally to get an opening into the wall. And it -- I've seen it many times where that’s what happens. People get upset and they just want to give something a good swift kick. And the height of it is consistent with where that would be. And it certainly would explain how it got into the wall as quickly as it did. But generally speaking, that's what I would have expected happened in this particular case.”

Q. And then whoever did that would have put some other type of material in there to

A. Absolutely. Absolutely. Would have used whatever was readily available. They could have brought certain things with them and took them when they left. The sky’s the limit there. But --

Q. And that's your -- how you can attribute the damage that you see and --

A. Absolutely.iii

The investigator apparently believed that the person kicking a hole through the wooden wall got very lucky and was able to avoid encountering a stud behind the wall. When what really happened is that this section of wall became involved late in the fire after fall down occurred inside the wall. The investigator also tightly embraced negative corpus methodology in this testimony:

One by one we try to address every single potential accidental ignition source in our area of origin. And once we are able to eliminate them all, there's absolutely nothing left that could have caused the fire except human involvement and that's the only way that we're able to do that.”iv

It was exactly this kind of flawed and unscientific thinking that caused the NFPA Technical Committee on Fire Investigations to disparage the use of negative corpus methodology in the 2011 edition of NFPA 921.

Figure 9.4 (f) Base of the wall under the actual point of origin in the electric service panel mounted above.

Failing to find a competent ignition source at the proposed origin should have led the investigators to look for a different origin but these investigators apparently wanted to find an incendiary cause regardless of the facts. Ten feet away from the proposed origin, there was a very similar fire pattern behind the gypsum drywall, shown in Figure 9.4 (f), which coincidentally was directly below the electric service distribution panel, seen in Figure 9.4 (g). Here was a fire pattern that was nearly identical to the fire pattern selected as the origin, but with an actual potential ignition source above it.

Unfortunately, the insurance company’s retained electrical engineer failed to recognize the significance of the evidence in this panel, from which the cover had been

removed prior to the fire. The panel exhibited numerous arcing events inside, including two holes arced through the side of the steel case, as shown in Figure 9.4 (h).

Figure 9.4 (g) Electric service distribution panel in place

after removal of some drywall. The cover had been removed

from this panel before the fire to deal with electrical problems.

When Mr. Svare conducted his re-examination of the panel and other electrical evidence, he noted melted lugs, illegal wiring techniques, and numerous arcs, as shown in figure 9.4 (i). This panel was clearly the origin of the fire, and the first electrical engineer’s failure to see that would prove costly later.

The fact that the state’s and the insurance company’s investigators determined both the wrong origin and the wrong cause had already been litigated, so it was new evidence that won the day for Mr. Awe. There was an evidentiary hearing held on September 17, 2012 at which both the author and Mr. Svare testified. The Judge, Richard Wright, found that the changes in NFPA 921 between the time of the original trial in 2007 and 2011 constituted “new evidence.” He even noted the similarities between arson cases and

shaken baby” cases, where the verdict is utterly dependent on the investigator’s opinion as to whether a crime had even occurred. He wrote:

Figure 9.4 (h) Electric service distribution panel, showing two circled areas

where arcing through the steel case had occurred.

Here the State's other evidence as to a set fire origin, to say the least, was weak; so this is not harmless. There was no direct evidence of the defendant's guilt. Circumstantial evidence of motive (financial hardship), and removal of keepsakes prior to the fire (just as likely consumed by the fire) added little. There was much argument over evidence regarding whether the electrical breaker box was involved and the location of the origin, and those arguments are not resolved by the new evidence. But the case was utterly dependent on the expert's opinion that this was a “set fire.” Had the jury learned that the State's experts had used a methodology now disapproved by a mainstream arson investigation association, there is a reasonable probability it would have had reasonable doubt as to the defendant's guilt. The result would have probably been different.

No one of the State's experts could determine a cause of the fire; they had theories as to potentialities, but no cause. From that they concluded, by elimination of hypothetical accidental causes, that the fire was not accidental. Without that opinion, there could hardly have been proof beyond a reasonable doubt. If it could have been accidental, it would have taken some strong evidence to point the finger at the defendant, and there was no such evidence.

This is not the fault of the State's arson investigators, who were trained in the flawed methodology. It is the result of the maturation of the arson investigation field, a gradual process of taking a second look at the negative corpus thinking.v

Figure 9.4 (i) Electric service distribution panel with orange zip ties

showing the locations of the arcing events that occurred inside the panel.

Judge Wright ordered a new trial and instructed that Joseph Awe be released from custody. The state declined to prosecute, because they would again have to rely on negative corpus methodology. Mr. Awe was not content, however. He filed a lawsuit against his insurance carrier for promoting his prosecution as well as for the policy limits they owed. He also sued the electrical engineer for failing to properly interpret the evidence of electrical activity in the service panel. Both of these cases were settled before trial.

This is not the fault of the State's arson investigators, who were trained in the flawed methodology. It is the result of the maturation of the arson investigation field, a gradual process of taking a second look at the negative corpus thinking.

Hon. Richard O. Wright, 2013

Error Analysis

The electrical engineer misinterpreted critical data by attributing electrical arcing inside the panel to an external attack, and erroneously excluded the panel as the ignition source. He also relied on the fire investigators for their origin determination, and was able to use circular logic to eliminate the panel because it was outside their area of origin.

The investigators also misinterpreted critical data. The obvious fall down at their false origin was interpreted as the fire moving from the outside to the inside, or alternatively, starting inside the wall as a result of a deliberate act. They interpreted the absence of the burned poster as evidence of prior removal, when any reasonable hypothesis about how the fire progressed would not have resulted in any evidence of the poster being left behind. Further, they did not consider that photographs are no longer the irreplaceable sentimental items that they were before digital photography became the norm.

The use of the negative corpus methodology is a form of misinterpreting critical data. The absence of any fuel or ignition source at the proposed origin was misinterpreted as evidence of an incendiary fire, rather than as evidence that another origin was possible. The investigators failed to give adequate consideration to the fire pattern directly beneath the electric service distribution panel, and went to great lengths to explain (incorrectly) that if the panel were at the origin, it would have exhibited much more damage.

The investigators also ignored inconsistent data. Mr. Awe was demonstrably 40 minutes away when the fire was discovered. He had an airtight alibi. The state and insurance company investigators “explained this away” by alleging, without evidence, that Mr. Awe had hired someone to set the fire. The reconstruction of the shelf provided inconsistent data, but instead of considering that data, the investigators simply moved the alleged origin to the interior of the wall and added the improbable and evidence-free explanation that some kicked or cut a hole in the wall. They also ignored the finding of $900 in cash in the till. If Mr. Awe had a financial motive, why would he burn up so much money? Maybe $100 or $200 could have been left in the till to mislead the investigators, but $900 certainly seems excessive. Attributing inconsistent data to a diabolical plot by the arsonist to mislead investigators allows every fire to be called arson.

This combination of errors led to an innocent man being robbed of three years of his life and living under a cloud of suspicion for six. This case should never have been brought to trial.


This is the first (so far the only) case known to the author in which a change in the science of fire investigation was found to constitute “new evidence,” that triggered the overturning of an arson conviction. This case also points out that photographs are no longer the irreplaceable items that they once were. Finally, the lawsuit against the electrical engineer demonstrates that errors can have real consequences for negligent fire investigators as well as for the falsely accused.

i WI v Joseph Awe, No. 07-CF-54

ii WI v Joseph Awe, No. 07-CF-54, Testimony of James Sielehr, December 18, 2007, Trial Transcript at page 192/258

iii Id. At 119/234.

iv Id. At 182/258

v Wright, R., (2013) State v Awe, Decision and Order

A personal note: After the case was over, appellate counsel sent me a copy of the allegedly missing photograph, which hangs on my office wall today.

If you are not on the mailing list, please click on “Join the Mailing List” to automatically receive the monthly newsletter.

Thank you for your interest. If you have any suggestions for future articles please email me at scientific.fire@yahoo.com

Subscribe to Our Newsletter

firstname :