Court Grants Motion to Exclude Insurer's Computer Fire Modeling Expert's Testimony

PROPERTY – HOMEOWNERS – FIRE INVESTIGATION – COMPUTER FIRE MODELING – ADMISSIBILITY OF EXPERT TESTIMONY – FRYE HEARING
Santos v. State Farm Fire & Cas. Co.
(Sup. Ct., Nassau Co., decided 6/28/2010)


Plaintiff moved pursuant to Frye v. United States, 293 F. 1013 (DC Cir. Ct. Apps.1923) to exclude the testimony of State Farm's defense computer fire modeling expert witness, Dr. Jozef Urbas.  At the time the motion was made, plaintiff contended that Dr. Urbas utilized the computer fire modeling to determine the origin and cause of a fire that occurred at plaintiff's premises in Bethpage, New York, on January 20, 2006. The court ordered a Frye Hearing.

Pursuant to the Frye test, expert testimony based on scientific principles or procedures is admissible only after a principle or procedure has gained general acceptance in its specified field (People v Wesley, 83 NY2d 417, 422 [1994]). A particular procedure need not be unanimously indorsed by the scientific community but must be generally accepted as reliable (Id. at 423). The Frye test emphasizes "'counting scientists' votes," rather than verifying the soundness of a scientific conclusion (Parker v Mobil Oil Corp., 7 NY3d 434, 447 [2006]).

In 1993 the United States Supreme Court decided Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 US 579 which, in federal courts, displaced the Frye general acceptance standard based upon the Federal Rules of Evidence. The Frye general acceptance test, however, continues to be the standard for determining reliability and admissibility of expert testimony in New York (see People v. Wesley, 83 NY2d at 433).

The burden of proof is on the party challenging the evidence to make a prima facie showing that it is a novel theory which is not generally accepted (Matter of Seventh Jud. Dist. Asbestos Litig., 9 Misc 3d 306, 311-312 [Sup. Ct. Wayne Co. 2005]). The burden then shifts to the proponent of the evidence to show by a fair preponderance of the credible evidence that there is sufficient general acceptance of its reliability (Id.).

Eugene J. West, retired investigator for the New York City Bureau of Fire Investigation (the Fire Marshal's Office), testified for the plaintiff that in his capacity as a fire investigator he would get many requests to utilize computer fire modeling but would decline to use it as part of the official investigation, stating that: "We can only speculate as to what was the composition or the exact construction of a room or the type of materials that were used, especially if those things are no longer available to us."

Mr. West stated that computer fire modeling was never generally accepted as an investigative tool by the New York City Fire Department.  It was his opinion that computer fire modeling is not generally accepted in the fire investigative community and cannot be used to determine the cause of a fire.  Although computer fire modeling was used in the World Trade Center investigation, it was used for illustrative purposes.  Mr. West explained that there is a caveat by the National Fire Protection Association ("NFPA") that the program is essentially as good as the information put into it.

In support of the introduction of the computer fire modeling evidence, Dr. Urbas testified the NFPA endorses the use of fire modeling in fire dynamics in the use of fire investigation.  Dr Urbas further testified that the underlying equations and laws of physics have been generally accepted in the fire science community and that they have been generally accepted as reliable in computer fire modeling.

State Farm's counsel explained that Dr. Urbas was not coming into court to state the cause and origin of the fire but rather to apply the computer dynamics to see how the fire would spread.  Dr. Urbas testified that the results of the fire modeling established that there was a time line that matched a particular origin of the fire, that the damage in the building corresponded to the results of the modeling and that the determination of fire dynamics in that particular theory [the time line] is generally accepted for that purpose.  The computer fire modeling essentially verified the hypothesis as to the ignition source or cause of the fire.  Dr. Urbas "never said it's accepted for determining the origin of the fire" and acknowledged that although "[i]t can help determining the cause", " it cannot be the sole method of determining the cause" of a fire.

After hearing the testimony of both parties' experts, Nassau County Supreme Court Justice Thomas Phelan granted plaintiff's motion to exclude the computer fire modeling testimony:

The NFPA Users Manual for 921 states: "To conduct valid modeling and testing it is important that the investigator gather data that is as accurate and complete as possible."  Dr. Urbas agreed that the concept "garbage-in-garbage-out" was applicable (p. 73).  Here, the input was based upon regulatory agency tables [for furniture, floors, walls, etc.]*, measurements taken by Dr. Urbas, his inspection of the damage and his reliance upon information received from fire investigators from the insurance company.  Dr. Urbas never spoke with the homeowner, the Nassau County fire officials or local firefighters.  Dr. Urbas testified that he was unaware that there were paint thinners and solvents in the area where the fire started and that such knowledge would have thrown off the entire calculation (p. 75).

Fire modeling carries with it a 15 to 20 percent margin for error assuming all conditions are correct but could be as high as 80 percent depending upon the real conditions (pp. 76-77).  Dr. Urbas acknowledged that there could be a difference between the material represented in a table and the actual material at the fire scene (p. 83). "If the input in the model is correct then the output is correct" (p. 96).  Dr. Urbas' testimony indicated that incorrect input could lead to inaccurate results.

Bearing in mind that the court's role as gatekeeper is "counting scientists," the court finds that defendant did not present sufficient evidence to establish that there is a consensus in the fire investigation community that computer fire modeling is generally accepted as reliable. "The long-recognized rule of Frye v. United States (supra) is that expert testimony based on scientific procedures is admissible but only after a principle or procedure has gained general acceptance' in its specified field" (People v. Wesley, 83 NY2d at 422).  Here the specified field is fire investigation. The issue before the trier of fact in this particular lawsuit involves when and how the fire started.  Defendant failed to meet its burden of proving that its expert's use of computer fire modeling was generally accepted in the fire investigation community (Cumberbatch v. Blanchette, 35 AD3d 341, 342 [2d Dept. 2006]). Although defendant's expert may support a case for the acceptance of computer fire modeling in the regulatory/design community, it does not support a conclusion that it is generally accepted in the fire investigation community. 

It is important to understand that Justice Phelan did not hold that computer fire modeling is scientifically unreliable and never admissible in a civil arson case; he merely held that in this particular case, State Farm did not submit sufficient evidence to convince him that computer fire modeling was generally accepted in the fire investigation community.

And as late as the mid-nineteenth century, some still believed that the earth was flat.