Opinion
C.A. No. 02C-12-002(THG).
Submitted: October 7, 2004.
Decided: November 15, 2004.
Deborah Crinigan, Esq., White Williams, Philadelphia, PA.
Sean A. Dolan, Esq., Marshall, Dennehey, Coleman, Warner Goggin, Wilmington, DE.
Thomas Bouchelle, Esq., Bouchelle Palmer, Newark, DE.
Philip Edwards, Esq., Murphy, Spadaro Landon, Wilmington, DE.
Dear Counsel:
Defendant, Rheem Manufacturing Company ("Rheem"), has filed a Daubert motion questioning the reliability of expert opinions on which the Plaintiffs seek to impose liability against it. The Defendant objects to the testimony by the Plaintiffs' experts Thomas Schneiders ("Schneiders") and Glenn Frederick ("Frederick") as to their theories of fire causation. The Defendant contends that the experts' opinion that a design defect of the Rheem water heater contributed to the fire is unreliable. The parties have submitted legal memoranda including reports, affidavits, depositions and other documentary materials which may be relevant to industry standards concerning hot water heaters. Oral arguments for this motion took place on October 7, 2004. The trial is currently scheduled for December 6, 2004. This is the Court's decision granting Rheem's motion in limine.
Mr. Frederick's opinion was proffered for the purpose of fire causation only. Plaintiffs' counsel acknowledge during the Frederick deposition that she was not "proffering this witness as an expert or to give testimony as to the origin area." (Frederick Depo. at 42-43). The origin portion of the opinion was addressed by Mr. Schneiders.
Michael and Sherry Podrasky entered into a construction contract for the building of their new home on January 29, 1999. On June 25, they purchased and ordered the installment of a 50-gallon water heater from Sharp Energy. A short time thereafter, the water heater was delivered and installed to the Podrasky home, which was still under construction. During its initial installation, Mr. Podrasky noticed wires on top of the water heater, and became concerned when the service technician burned his hand on the unit. Based on his observations and comments made by the installation technician, Mr. Podrasky complained to Sharp Energy manager, Bennett Prettyman, who assured him the water heater was safe and refused to replace it. During its final installation and startup, Mr. Podrasky observed yellow flames coming from the heater. The service technician installing the unit assured him again that there was no need to be concerned.
The Podraskys moved into their new home in December 1999. From the date they moved in, they experienced an "intermittent problem with the temperature of the heated domestic water for the structure." The Podraskys lived in the home for approximately one year before the fire occurred. Sometime during the evening of December 3, 2000, the Podraskys experienced problems with the gas furnace in their home. The next morning, Mrs. Podrasky ran out of hot water during her shower. Later that morning, Mr. Podrasky also noticed a shortage of hot water during his shower. Mr. Podrasky called Sharp Energy to request a service visit to their home. Later that day, Ms. Podrasky was called from work to meet the technician at her home. When she opened up the door to the garage, where the water heater was located, she noticed smoke and water coming out of the structure. Mr. Robinson, a Sharp Energy employee and Mrs. Podrasky demonstrated quick thinking by shutting off the propane gas flow to the home and alerting emergency officials. Mr. Robinson attempted to fight the fire until the fire fighters arrived. After the fire was extinguished, an investigation by Fire Marshal Lee determined that the blaze originated in the area of the hot water heater and was caused by the proximity of the wooden platform to the water heater above it.
(Pl.'s Memo. of Law of 9/17/04).
Shortly after the fire, the Podraskys retained the Dove Associates firm to conduct an independent investigation into the cause of the fire. Based on the results of this investigation and the damages they incurred, the Plaintiffs' brought this complaint alleging "damage and destruction to their real and personal property, extra expenses, additional living expenses, and incidental and consequential damages in an amount in excess of One Hundred Thirty-Four Thousand Dollars ($134,000.00)."
(Pl.'s Praec. at 24).
Thomas Schneiders and Glenn Frederick were assigned by Dove to conduct a fire origin and cause investigation. Thomas Schneiders has extensive experience in the fire fighting and investigation fields. He testified that he has fought and investigated many fires involving gas-fired appliances including water heaters. Schneiders also testified that, over the course of his career, he has installed approximately 100 water heaters as personal favors for friends. His training for the Philadelphia Fire Academy and Department included several training sessions concerning gas-fired appliances and how to best approach to handling emergencies involving these types of fires. Schneiders' involvement with fire fighting and investigations, instruction as to fire origins and causes, and consulting experience in the field evidence his experience in determining the origin and cause of fires. But, the Court feels it is imperative to note that Schneiders' installation experience and training at the Fire Academy does not equate to sufficient specific knowledge, training or experience in the area of water heater design and operation.
(Schneiders Aff. ¶ 4).
(Frederick Depo. at 7).
The other expert, Glenn Frederick, was assigned to the Podrasky case for his engineering experience. Frederick obtained a Bachelor of Science degree in Engineering from Drexel University. He is a licensed engineer in four (4) states and has thirty years (30) of experience as an engineer. His education involved studying the "prerequisites of all the things that come together on the appliance," such as principles of heat transfer, combustion, and compressible fluid flow, but did not include a specific focus on the operation of particular gas-fired appliances. His education included "training in college with regard to the manufacture and design of gas-fired water heaters," but his experience after college with regard to gas-fired appliances focused mostly on installation and servicing. Frederick testified that some of his career experiences involved the design and modification of heating systems and other mechanical systems. Frederick was proffered by the Plaintiffs to give an opinion on fire causation. In his deposition, Frederick explained that he had no official training in determining the cause of fires, but that during his career he has been consulted on approximately one thousand (1000) cause and origin fire scenarios. His knowledge of fire origins and causes developed entirely from on the job training.
See id. at 9-10.
See id. at 18-19, 21.
See id. at 37-38.
As part of the investigation, Mr. Schneiders visited and examined the fire scene on December 7, 2000. He took photographs of the scene and conducted interviews with the Podraskys about the water heater's history in the home, the morning of the fire, and Mrs. Podrasky's observations of the actual fire scene. Schneiders also spoke with Fire Marshal Lee, who conducted the investigation immediately following the fire. Schneiders concluded that the wooden stand beneath the water heater was the origin of the fire by examining the damage, burn patterns and other evidence at the scene. According to accepted fire investigation principles, to determine the cause of a fire, "the investigator has to identify the source and form of heat ignition, the first material ignited, and the ignition factor." Schneiders identified the water heater as the only heat producing source in the area of the fire, and ruled out all other potential ignition sources.
(Pl.'s Memo. of Law of 9/17/2004 citing NFPA 921, Section 12.1-5).
Mr. Schneiders' investigation and Mr. Frederick's review of the photographs from the fire scene and examination of the water heater yielded a hypothesis. They believed that "the masonry material found in the combustion induction fan caused the abnormal operation of the water heater and a discharge of heat energy from the bottom of the unit sufficient to ignite the wooden stand." Frederick testified in his deposition that "as a good engineer and scientist, I have to develop a test to study that hypothesis, to determine the viability of that hypothesis. This is what I did." Frederick designed a test "to solve a specific hypothesis" by attempting to simulate what he and Schneiders believed occurred within the Podrasky's water heater.
Hypothesis is defined as "a proposition tentatively assumed in order to draw out its logical or empirical consequences and test its consistency with facts that are known or may be determined." See Merriam-Webster Medical Dictionary (2002).
(Pl.'s Memo. of Law of 9/17/2004 at 6).
(Frederick Depo. at 85). During oral argument, Plaintiffs' counsel suggested that the test was an unnecessary component of the fire causation opinion. She argued that she wished the experts had never conducted the test. However, Frederick acknowledged in this response that a good scientist recognizes the need to test a hypothesis to ensure its viability. Without testing, a hypothesis remains, as it is defined, a "proposition tentatively assumed." See Merriam-Webster Medical Dictionary (2002). Unfortunately, the testing of this hypothesis did not yield supportive data. The test did not isolate the malfunctioning fan and safety setting as the primary variables in the generation of heat from the water. The flawed testing methodology resulted in an opinion that assumes rather than proves that the heat generated by the malfunctioning water heater was sufficient to ignite a wooden stand below it.
(Frederick Depo. at 85). Frederick testified that he "designed a test to test the hypothesis of how this fire started, based on all the visual information we had, based on all the evidentiary information we had." (Frederick Depo. at 65).
The test was conducted in July 2001 at the Dove Associates facility. Using an exemplar water heater, identical to the Podraskys' model, Frederick and Schneiders attempted to simulate the obstruction of the fan and the reduced air pressure within the unit. They conducted two preliminary tests. The first was done to ensure that the combustion induction fan and its motor could be controlled as the test required. The second test was conducted to "determine the characteristics of the vent system and the vacuum switch."
(Dove Assoc. Rpt. at 10).
The third and final test completed by Schneiders and Frederick is the real focus of this motion. It was designed to simulate the reduced air pressure in the combustion chamber of the unit to determine how much heat exited the unit functioning under abnormal conditions. The test was conducted outdoors for a period of twenty (20) minutes. During the testing period, a thermocouple placed near the leg of the exemplar water heater, on the surface of the wood, measured air temperatures as high as 193 degrees Fahrenheit. Frederick also reported seeing flames exit the bottom of the chamber, however, his inconsistent reporting of this occurrence causes the Court great concern, which will be discussed below. The experts also noted the existence of significant conduction energy, evidenced by the warping of the wood near the support leg of the water heater.
The report generated by Schneiders and Frederick asserts that heat energy "convecting, conducting, and radiating from the bottom of the exemplar water heater was sufficient to result in ignition of the wooden support surface over a period of time and operation." The experts concluded that the accidental fire was caused by a long term energy discharge from the water heater, which was caused by the abnormal operation of the unit's combustion fan. The experts opined that the abnormal operation of the fan was a result of drywall found in the fan, which indicated poor workmanship in the unit's installation. They claim that proper installation, servicing and maintenance of the water heater should have detected the presence of the drywall and removed it. The Plaintiffs' experts further noted that the safety controls included on the unit should have detected the reduction in air flow, and that the failure to do so is a design defect. The testing conditions and conclusions were summarized in a Dove Associates report issued on August 23, 2001. Frederick later testified in his deposition that the report contained all the information he relied upon in making his determination as to fire causation.
See id. at 11-12.
(Frederick Depo. at 103).
On July 29, 2004, Defendant Rheem submitted a Motion in Limine seeking to exclude these opinions of fire causation at trial. First, the Defendant objected to the experts' reliance on pyrolysis as a factor in the fire, arguing that pyrolysis has been rejected by the 10th Circuit as an unproven basis of ignition. Second, according to the Defendant, the experts proffered by the Plaintiffs are not qualified to opine on the design or operation of water heaters.
Third, the Defendant claims that the experts' test cannot withstand a Daubert review because it was created by Frederick, and was neither subjected to peer review nor supported by the engineering community. The Defendant claimed that the proper test for evaluating the flooring temperatures for water heaters was set forth by the American National Standards Institute ("ANSI"). The relevant ANSI test places a properly functioning water heater inside a wooden box (to simulate a closet) at a Nationally Recognized Testing Laboratory and measures the temperature of the flooring below. The Rheem water heater at issue in this case had been subjected to the applicable testing done by ANSI, and had passed all the standards required to be ANSI certified.
(Def.'s Mot. to Exclude of 7/29/2004, at 10).
According to its website, the American National Standards Institute is "a private, non-profit organization (501(c)3) that administers and coordinates the U.S. voluntary standardization and conformity assessment system." See American National Standards Institute website, available at http://www.ansi.org/about_ansi/overview/overview.aspx?menuid=1, last verified 11/09/2004. The purpose of ANSI is to set nationwide standards for the testing of household appliances and other products. Plaintiffs' counsel has suggested at various points throughout this proceeding, that the ANSI standards are not the applicable standards to apply to the testing of this water heater. The Court acknowledges that there are separate standards for fire origin and cause investigations, set by the NFPA, and other standardizing agencies besides ANSI, such as the ASTM, which sets standards for household products as well. The Court is satisfied, however, with classifying ANSI as an agency that sets the nationally recognized standards within the water heater industry. In his deposition, Frederick acknowledged that ANSI "set[s] testing standards that [water heaters] must meet . . ." (Frederick Depo. at 53).
(ANSI/CSA Certificate of Compliance of 4/29/1999).
The Defendant contends that the experts' conclusions are not in line with the accepted standards of the water heater industry if "the temperatures created under the ANSI test were found to be acceptable . . . [whereas] Mr. Frederick's test probably produced lower flooring temperatures . . . but Mr. Frederick somehow determined the heat produced by the Rheem water heater during his test was unacceptable and dangerous." Frederick, himself, acknowledged the possibility that a water heater could be certified under ANSI standards, but fail his test, which produced less heat because of its placement outdoors.
(Def.'s Mot. to Exclude of 7/29/2004, at 6).
(Frederick Depo. at 90).
Rheem insists that the applicable ANSI test was the proper evaluation of whether the water heater emitted dangerous heat to the floor below, and that its certification demonstrates that it does not. The Defendant contends that the experts' basis of opinion cannot be accepted because the test they employed "has no basis in engineering literature, has never been subject to peer review and has not been recognized in the engineering community."
See id. at 10.
Fourth, the Defendant claims that the experts' test cannot be properly applied to the facts of this case because the test was terminated prematurely, before achieving ignition. Because the experts failed to demonstrate that the wooden stand would eventually ignite from the heat produced by the water heater, the Defendants argue that the experts have relied improperly on theories of pyrolysis to provide the ignition source. The theory of long term, low temperature ignition of wood has been rejected by the 10th Circuit as an unreliable hypothesis without sufficient testing. Moreover, even accepting pyrolysis as valid, the Defendants contend that the experts failed to demonstrate the process of pyrolysis in their test, arguing that the experts ended their test before any indications of ignition ever arose.
See Farmers Truck Ins. v. MagneTek, Inc., C.A. No. 00-RB-2218 (CBS) (D. Col. 2002).
Fifth, Rheem alleges that the test results cannot be applied to the case at hand because the experts used a different type of wood than was present in the Podrasky case. Moreover, they did not measure the temperature of the actual wood beneath the water heater, and instead measured the temperature of the surrounding air. Arguing that the test only produced temperatures of 193 degrees Fahrenheit, compared to the 450-470 degrees typically required for the ignition of wood, the Defendant contends that the expert opinion relies on pyrolysis for ignition.
The Plaintiffs responded by explaining the findings of the experts more precisely. First, they claim that the experts' causation opinion was not solely based on pyrolysis, but also depended on the presence of radiant energy emanating from the water heater. They explained that it was a combination of the convection, conduction and radiant heat that ignited the wood below the stand. Furthermore, they explained that the "test [was] designed to show that the Rheem water heater was capable of producing enough heat to be an ignition source . . ." Second, the Plaintiffs argue that both Schneiders and Frederick are qualified to be considered experts in fire causation. Mr. Schneiders' qualification comes from his extensive experience as a firefighter and investigator. Mr. Frederick's expertise is evidenced in his engineering background, including his formal education and thirty years of experience in the installation, repair and design of water heaters.
(Pl.'s Resp. In Opp. To Def.'s Motion of 9/17/2004, at 1).
See id.
See id. at 2.
The Plaintiffs argue that the test utilized by the experts was appropriate to test their particular hypothesis and properly conducted. They argue that the test was not designed to prove that the heater, in malfunctioning condition, could ignite the wood below it, but instead to measure the level of heat that it emitted. Because the test was not conducted to prove ignition, the Plaintiffs' claim that it was not necessary to assimilate the exact conditions present in the Podrasky garage. They also argue that any variations caused by the different testing conditions strengthens their position since the presence of air drafts, as opposed to placement inside a wooden box, would result in lower temperature readings than the water heater would produce indoors. Finally, they object to the Defendant's characterization of the ANSI Heat-Roll Out test as the appropriate test, since that test is based on a properly functioning water heater. Instead, the Plaintiffs claim that their test was a modified version of the Heat Roll-Out test, specifically designed to measure the heat roll-out of an improperly functioning water heater. The Plaintiffs claim that their experts' test was properly conducted and terminated once the test yielded sufficient evidence of excessive heat exiting the base of the heater.
The rule governing the admissibility of expert opinions in Delaware courts is Delaware Rule of Evidence 702. It provides that "[i]f scientific, technical or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training or education may testify thereto in the form of an opinion or otherwise . . ." When interpreting this rule, which is identical to Federal Rule 702, Delaware courts have adhered to the guidelines set forth by the United States Supreme Court.
See D.R.E. 702.
See M.G. Bancorporation, Inc., 737 A.2d 513, 521 (2003).
In 1993, the United States Supreme Court changed the standards of admissibility for scientific opinions. In Daubert v. Merrell Dow Pharmaceuticals, the Court overruled the previous approach to expert opinions, which relied on the relevant scientific community's general acceptance of the methodology underlying the testimony. In its place, the Court adopted a standard which gave the trial judge the responsibility of being a gatekeeper for the scientific evidence admitted in a trial. The Court interpreted Federal Rule 702 as requiring a trial judge to ensure that all scientific expert testimony is relevant and reliable evidence. In a later case, the United States Supreme Court extended this gatekeeping function to include the governing of all expert testimony, not just scientific evidence. Delaware courts have closely followed the Daubert and Kumho decisions, as well as the guidance these decisions lend as to how to determine the reliability of potential testimony.
See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589 (1993).
See Frye v. United States, 293 F. 1013 (1923).
See Daubert, 509 U.S. at 597.
See id. at 589.
See Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 143 (1999).
See M.G. Bancorporation, 737 A.2d at 522.
To properly admit scientific expert testimony, a trial judge must ensure that the inclusion of the testimony will assist the trier of fact. Next, "an inference or assertion must be derived by the scientific method. Proposed testimony must be supported by appropriate validation — i.e., "good grounds," based on what is known. In short, the requirement that an expert's testimony pertain to "scientific knowledge" establishes a standard of evidentiary reliability." However, trial judges may not always have the requisite expertise to judge the reliability of scientific testimony. For this reason, the Daubert court set forth various factors that a trial court may consider when deciding whether certain evidence is sufficiently reliable to admit into evidence. The four-part test in Daubert instructs a trial judge to determine whether the proffered testimony:
See Daubert, at 591.
See id. at 593-94.
1. has been tested;
2. has been subjected to peer review or publication;
3. is a result of a technique with a known or potential rate of error; and
4. is generally accepted within the scientific community.
See id.
These considerations are not absolute, and the trial judge is left with wide discretion to consider other relevant factors.
See id. at 592-93; Goodridge v. Hyster Co., 845 A.2d 498, 503 (Del. 2004).
In the present case, the Plaintiffs seek to present expert testimony as to the cause of the fire at the Podrasky home. The Plaintiffs opine that the Rheem gas water heater caused the fire, in part, due to the defective design of its safety controls. But, the cause of this fire may not be proven by relying on the principle of res ipsa loquitur. The Court is concerned that allowing the combined opinion of an experienced firefighter and an engineer, the jury will automatically assume that their assessment of this fire is reliable. This may also lead the jury to infer that their opinions concerning the alleged defect in the water heater are reliable. But the Court is responsible for not only verifying the reliability of the experts who testify, but also the reliability of that evidence they put forth. While there is some evidence that supports the argument that the fire began in the area of the hot water heater, that does not automatically result in finding Rheem responsible for the fire.
In 2004, the Delaware Supreme Court applied a five-part test to determine whether expert testimony was properly excluded from the trial phase. The Court was charged with deciding whether a safety expert could testify as to his opinion that any forklift without safety controls for backing up was unreasonably dangerous and defectively designed. When determining admissibility of the expert opinion, the Court considered:
See Goodridge, 845 A.2d at 503.
See id. at 500.
1. whether the witness qualifies as an expert by knowledge, skill, experience, training or education;
2. whether the evidence is reliable and relevant;
3. whether the opinion is based upon information reasonably relied upon by experts in the particular field;
4. whether the specialized knowledge being offered will assist the trier of fact to understand the evidence or determine a fact in issue; and
5. whether the expert testimony will create unfair prejudice, confuse or mislead the jury.
See id.
The Court held that the proffered expert lacked sufficient forklift design experience and engineering skills to opine on design aspects, and that his testimony would not assist the jury meaningfully. In reviewing the case for abuse of discretion, the Court instructed that
[t]he trial judge's inquiry should include whether the proffered expert and the purported "field of expertise" itself can produce an opinion that is sufficiently informed, testable and, in fact, verifiable on an issue to be determined at trial. Even though an expert may be qualified to opine within a recognized "field," that fact alone does not automatically guarantee reliable, and therefore admissible, testimony. It is critical that a trial judge be satisfied that any generalized conclusions are applicable to the particular facts of the case.
See id. at 503.
Applying Goodridge to the present case, and after reviewing the report and the experts' depositions, the Court is unpersuaded and concerned by the testing conditions, the reporting by the experts, the reliance on pyrolysis as applied to the facts of this case and thereby, the test results. All of these concerns, independently and collectively, necessitate the exclusion of the experts' testimony.
First, the test was conducted outside. The experts acknowledged in their report that the wind conditions on the day of testing were variable and had an effect on the draft pressure of the unit. "These variations in the draft pressure . . . resulted in the occasional stoppage and reestablishment of gas flow to the exemplar water heater's burner." They reported that "when propane gas flow was reestablished to the burner after an outage the initial flame extended from the bottom of the water heater . . ." It seems clear to the Court that the variations in draft pressure reported by Frederick and Schneiders could have had a substantial effect on the testing conditions and results. Fundamental science principles require that the number of variables in an experiment be limited. Otherwise, its results cannot be reasonably attributed to the target variable being tested. Here, if the wind caused the burner to extinguish and ignite as the experts reported, then the flames and heat produced by the exemplar test could have been a result of the wind rather than the reduced air pressure within the combustion chamber. The Plaintiffs' theory of fire causation partially depends on the fact that the water heater was cycling on and off due to the irregular operation of the combustion fan. If the air drafts contributed to this cycling and consequential heat rollout, then the test did not properly isolate the reduction in air flow as the cause of the heat rollout. The apparent and obvious differences between the actual fire scenario, which occurred indoors without the presence of significant air drafts in December, and the testing conditions, which occurred outside exposed to variable air drafts in July, are too great to consider this test reliable and relevant.
(Dove Assoc. Rpt. at 11).
See id.
In their report, Schneiders and Frederick report that the air draft variations resulted in the "occasional stoppage and reestablishment of gas flow to the exemplar water heater's burner." (Dove Assoc. Rpt. at 11). They also reported that "when propane gas flow was reestablished to the burner after an outage the initial flame extended from the bottom of the water heater." See id. In a later affidavit, Frederick reported that the "fired test showed large amounts of heat energy released from the base of the exemplar water heater when ignition of the gas burner occurred. The test also showed that the gas burner would ignite then extinguish, then ignite again, in rapid succession over and over." (Frederick Aff. ¶ 15). The Court takes two issues with this inconsistency. First, the reporting as to how often the flames were exiting the bottom of the water heater is inconsistent. The occurrence goes from "occasional" in the report to "rapid" in the affidavit. Second, the emission of flames from the bottom of the heater may have been caused by the presence of air drafts which would not have been present in the Podrasky home. The reporting and the apparent flaws in testing conditions are fatal to the reliability of this test.
The Plaintiffs have suggested that the fact that their test was subject to natural air drafts and outdoor temperatures actually weighs in their favor. They contend that the presence of air drafts indicates that their temperature readings were actually lower than the actual heat energy emitted by the water heater because of the cooling effect of the air drafts. Unfortunately, the wind also resulted in the interference of gas flow, causing the water heater to ignite and extinguish repeatedly. This destroyed the reliability of their test. The Court understands that reliable and relevant test results may be produced outside of a Nationally Recognized Testing Laboratory. Reliable test results, however, require predictable and controlled testing conditions. A more reliable test would have mimicked the conditions of a water heater's everyday use. Water heaters are typically housed indoors. The `backyard test' that Schneiders and Frederick conducted cannot be depended upon to demonstrate the heat energy capable of emission from the water heater, because the test left too many variables uncontrolled.
I also note that the experts used the above conclusion to opine that "the failure of the water heater's safety controls to properly and safely react to the reduced flow of combustion air resulting from the presence of foreign masonry material in the induction fan is a failure in the design of this appliance." However, the experts' opinion is void of any discussion of design and engineering practices in the water heater industry. There is no proper or complete analysis of this conclusion or opinion.
(Dove Assoc. Rpt. at 13).
The Delaware Supreme Court ruled in Goodridge v. Hyster that an occupational safety expert could not testify as to his opinion that forklifts without backup safety devices were unreasonably dangerous and defectively designed. Basing its decision on the proposed expert's lack of engineering and design courses, as well as the design experience to testify as to design defect, the Court found that there was an "analytical disconnect between [the expert's] experience in occupational safety and his ability to give a reliable opinion about whether an injury was caused by an unreasonably dangerous and defective forklift." The case at hand presents the same problem. Unlike the expert in Goodridge, Frederick possesses a degree in engineering and has worked with gas-fired appliances throughout his career. However, the lack of design experience and an absence of experience in the manufacture of gas water heaters disqualifies him as an expert on an issue of design defect.
See Goodridge, 845 A.2d at 504.
See id.
Neither Schneiders nor Frederick has the requisite education, knowledge, training or experience to qualify as an expert on the design and safety characteristics of a water heater. Neither of them has been involved in the design or manufacture of water heaters beyond incidental installation and repairs. When asked whether his opinion about the safety switch's setting was supported or acknowledged in any form, Frederick responded that it was his "sole opinion." There has been no evidence presented to the Court that indicates that, since the experts' test or the Podrasky fire, the water heater industry has taken notice of a potential flaw in the setting of Rheem safety switches.
The Plaintiffs have submitted a report that notes the flame arrestor plate technology that has been incorporated into all water heaters over the past three years. The experts did not refer to this report in their opinion, but the Plaintiffs attempt to use it as evidence of Rheem's faulty design. The Plaintiffs argue that design changes were made by Rheem to address the heat rollout problems presented by this action. But this is not correct. This new technology may decrease the possibility of heat rollout, but it was not developed as a response to incidents involving decreased air flow through the combustion chamber. The "new line" of gas heaters was designed in response to new ANSI standards addressing the problem of flammable vapor ignition. This is not an issue in this case. The introduction of these materials for the purpose of buttressing the Plaintiffs' theory only demonstrates the weakness of the Plaintiffs' theory. The Plaintiffs have failed to present any evidence, other than their experts' unfounded opinions, that Rheem water heaters are set improperly to detect dangerous levels of air flow through the air chamber. Schneiders and Frederick will not be permitted to testify that the setting on the Rheem water heater constituted a design defect because they have not shown any reliable basis for this opinion, either in their own testing or commentary within the water heater industry.
A number of fires caused by gasoline fumes ignited by water heaters prompted the inclusion of flame arrestor plate technology on all water heaters produced after July 1, 2003.
Second, the inconsistencies found by the Court between the experts' report, depositions and affidavits raise an independent concern for reliability. A trial judge is charged with ensuring that the admission of expert testimony is both relevant and reliable. The testimony proffered by the Plaintiffs' experts has varied significantly during different stages of this proceeding. It appears to the Court that the experts have adjusted their opinions according to the arguments presented by the Defendant.
See Daubert, 509 U.S. at 597.
The most troubling of these shifting opinions concerns the fundamental basis of the experts' fire causation opinion. In Frederick's deposition, he explained that the presence of open flame during testing was not included in his report because it was unnecessary and not the basis of his opinion. His opinion instead seemed to rely heavily on the presence of conduction and convection and radiant heat energy which contributed to the process of pyrolysis. When asked why he conducted the test, Frederick responded "to see if enough heat would emanate from the bottom of the water heater to have caused pyrolysis to occur."
(Frederick Depo. at 147-48).
See id. at 123-24.
However, after the defense attacked the position for relying solely on pyrolysis as the ignition source, the experts responded with affidavits and a pre-trial submission that relied heavily on the flame factor. The report issued by the experts refers to the "occasional" reestablishment of gas flow to the unit's burner during the test, which resulted in flames extending from the bottom of the water heater. But the affidavits, filed after the attack on the fire causation opinion, noted that "the gas burner would ignite then extinguish, then ignite again, in rapid succession over and over."
(Dove Assoc. Rpt. at 10-11).
(Frederick Aff. ¶ 15, emphasis added).
Plaintiffs' counsel argued that their experts' opinions have not been adjusted, but instead that the selective questioning by defense counsel during the deposition phase ignored the contribution of open flames to the experts' ignition scenario. But the presence of flames during testing was actually covered thoroughly in Frederick's deposition, and Frederick explained its role in the fire causation opinion in his response. Frederick explained that flames were exiting the bottom of the water heater on restarts, but that he did not include that in his report because it was not the basis of his opinion. He explained that flames shooting out of the bottom of the water heater were not included in his report because "[i]t's not needed." When asked in his deposition if he included "within [his] written report all of those factors that [he] thought were important in terms of the cause of the fire", Frederick responded "I think I did." Frederick was given an opportunity to fully explain what contributing factors gave rise to this fire. At this point, the expert had an opportunity to clarify his opinion to include the importance of the open flame, but he did not. The deposition transcript clearly demonstrates that Frederick did not rely on the presence of flames to support his opinion before the Defendant objected to its reliance on pyrolysis. However, in a later affidavit, Frederick asserts that "[t]he radiant energy transfer and flame is the primary means of energy transfer with the presence of open flame to sustain combustion of the wooden stand." It seems peculiar to the Court that the Plaintiffs' position became focused on flames as the primary source of ignition after their previous position was attacked. The drastic change in opinion from a minimal and incidental mention of open flame to total reliance on its presence is a significant inconsistency that this Court cannot ignore.
(Frederick Depo. at 147).
See id. at 148.
(Frederick Depo. at 104).
(Frederick Aff. ¶ 16).
In an effort to convince the Court that its revised theory is properly founded in accepted scientific principles, the Plaintiffs included in its memorandum of law an exhibit containing the typical temperatures of adiabatic flames. Using this as support, the Plaintiff argued that their "test was conducted to determine the heat flow from the bottom of a malfunctioning water heater. The results confirmed that the radiant flames were at temperatures of 2521-3521 degrees Fahrenheit." The Court finds this assertion misleading and irrelevant to the facts of this case. The Plaintiffs' experts did not measure the temperatures of the flames exiting the water heater and have presented no other support for this assertion. They did measure a temperature of 193.6 degrees Fahrenheit based upon an instrument laying on top of the wood on which the heater was placed. But, the adiabatic flame article and table list the highest temperatures reached by flames under ideal circumstances in laboratory conditions. There is no support for the assertion that the flames exiting this water heater were in the same range as those contained in the table nor that a hot water heater could produce such high temperatures.
See Vytenis Babrauskas, Ignition Handbook, Adiabatic Flame Temperature Table, at 30-31.
Next, there are several inconsistencies contained in the record regarding whether Frederick used any accepted literature or guidelines in the design of his experiment. In his deposition, taken on April 2, 2004, Frederick stated that "[he] designed the test to solve a specific hypothesis." He then testified that there was no literature that supported his test. Frederick was also asked whether he referred to "any of the ANSI literature before conducting [his] experiment?" His responses were "Do I have knowledge of it; yes . . . Yes, I did. Did I specifically pull it out and look at it? No, I did not." Later, he explained the differences between the results of his test and an ANSI test, Frederick stated "they are different tests." When asked by Plaintiff's counsel whether he was "trying to duplicate what is done in the ANSI test," Frederick responded "No, not at all." Frederick's responses in his deposition clearly indicate that his test did not rely on any ANSI literature for support.
(Frederick Depo at 85).
See id.
See id. at 53.
See id. at 179.
The Court was therefore concerned when later submissions by the Plaintiffs' experts indicated that the test they employed was "modeled from the ANSI heat roll out safety shut off means test (ANSI Z21.10.1-1993) . . ." The affidavit also refers to several authorities on heat transfer and engineering principles that served as bases for the experiment's design. The sudden appearance of these authorities, offered as the foundation for this experiment, only after the raising of the Daubert motion, requires the Court to raise a skeptical eye to the record the experts have compiled.
(Pl.'s Resp. in Opp. to Mot. to Excl. of 9/17/2004, at 3; Frederick Aff. ¶ 7). The Plaintiffs have argued throughout this proceeding that the ANSI heat roll out test did not test their specific hypothesis because it tested a properly functioning water heater, whereas their test used a malfunctioning water heater. They used this explanation first to justify why they did not use the ANSI test at all, and then later used the same reasoning to explain why they "modified" the ANSI testing procedure. While the Court, again, takes issue with the shifting justifications, it is a moot point. After extensive review, it appears that both tests involve the reduction of air flow to a properly functioning water heater. The tests employed different means to reduce the air flow to the combustion chambers, but the result in both was to provide less air than needed for proper combustion and then to examine the consequences.
The ANSI heat roll out test reduces the air available for combustion to the lowest point without causing the unit to shut off. While operating at this reduced air flow, the water heater must demonstrate that "either the means functions to shut off the gas supply to main burner(s) or the measured temperature reaches and is maintained at a temperature not exceeding 210 F (116.5 degrees C) above room temperature." See ANSI Z21.10.1 § 2.8 (1993). The Rheem water heater model that was present in the Podraskys' home was tested in a Nationally Recognized Testing Laboratory and was ANSI certified. See ANSI/CSA Certificate of Compliance of 4/29/1999.
The Plaintiffs' test also reduces air flow, but in an irregular fashion. The Plaintiffs suggest that the "cycling" effect of the variant air flow may have resulted in the ignition firing repeatedly, which was not noted in the ANSI test. But the test did not conclusively prove that variant air flow from a malfunctioning fan caused repeated ignition and heat rollout. Instead, the experts admitted in their report that wind variables on the day of testing resulted in the occasional extinguishment of gas flow, and upon reestablishment, they observed flames and heat rollout. Because of the test's failure to isolate the cause of "cycling" and consequential heat rollout, it is unclear to the Court whether the experts have identified a valid problem or simply introduced an additional variable to the ANSI heat rollout test which results in "cycling." Moreover, the ANSI standards require that a water heater either shut off properly when experiencing reduced air flow or that its heat emissions not exceed 210 degrees above room temperature. See ANSI Z21.10.1 § 2.8 (1993). The highest temperature reading produced during this experiment was 193.6 degrees Fahrenheit.
(Frederick Aff. ¶ 7).
Daubert recommends that a Court consider the general acceptance, peer review and publication of a methodology before it is admitted into evidence. In its Motion to Exclude the opinions of Frederick and Schneiders, the Defendant highlighted the lack of any peer review, publication or general acceptance of the methodology and testing procedures used by Frederick in this experiment. The Plaintiffs' response to the Defendant's Motion, filed on September 17, 2004 and the experts' supplemental affidavits, taken September 14, 2004, present the existence of supportive authorities for the first time. Whether or not the test methodology was supported by the engineering and water heater industries weighs heavily in the decision to qualify this test as reliable. The inconsistencies with which the experts have presented support for their test methodology and conclusions forces the Court to rule the opinions based upon them unworthy of qualification. The simple fact is that there has been no validation, recognition or other commentary on this "unique" test, which was conducted in July 2001 and which apparently discovered a significant flaw in the design of a popular hot water.
See Daubert, 509 U.S. at 593-94.
(Def.'s Mot. to Exclude of 7/29/2004, at 10).
(Pl.'s Memo. of Law of 9/17/2004, at 7; Frederick Aff. ¶ 7).
The test proffered by the Plaintiffs' experts fails to pass the standards of relevance and reliability set forth by Daubert. To survive this motion, the Plaintiffs needed to demonstrate the reliability of the methodology employed by its experts. Conducting the test outside destroyed its applicability to the facts at hand. Without reliable test results to support his theory of fire causation, the theory itself becomes unreliable. Moreover, the inability of the experts to consistently report their observations, findings, and support for their opinions seriously detracts from their reliability.
Finally, the portions of the experts' opinions that are based on pyrolysis are not sufficiently grounded in science nor reasonably applicable to the facts of this case. As previously discussed, the Plaintiffs have distanced themselves considerably from the initial opinions by their experts that pyrolysis contributed significantly to the cause of this fire. In fact, Frederick explained in his initial deposition that the purpose of performing his "unique" test was "to see if enough heat would emanate from the bottom of the water heater to have caused pyrolysis to occur." When questioned about the temperature differentials between the normal ignition point of wood (between 400-450 degrees) and the temperatures recorded in his experiment, Frederick responded "[y]ou've got to throw the pyrolyzing effect into this. When you do that, the temperature of the wood, which it will auto ignite, drops. And it drops progressively the longer it's exposed. It will eventually, some texts say, get down to 150 degrees ignition temperature." This answer alone reveals that Frederick relied on theories of pyrolysis to explain how wood could "auto-ignite" at 193.6 degrees Fahrenheit. Schneiders explained the effect that pyrolysis had in their decision to stop the experiment before reaching ignition, saying "the pyrolysis effect could actually turn something into a piece of charcoal, where it actually becomes black in color before you have ignition . . . [w]e didn't get that wood to that temperature . . . but I know for a fact that wood would ignite at some time." Clearly, both experts relied on theories of pyrolysis to reach their conclusion.
(Frederick Depo. at 123-24).
See id. at 119.
Schneider Depo. at 141-42.
Neither the report nor the depositions clearly present radiant flames with temperatures exceeding 2000 degrees as the primary ignition factor. Furthermore, during their depositions, neither expert took the opportunity to explain how the flames exiting the water heater were the actual cause of ignition. Instead, they explained how pyrolysis worked to lower the ignition temperature of wood to as low as 150 degrees Fahrenheit, and that they stopped their experiment once they observed the beginning stages of pyrolysis. After the pyrolysis position was attacked in this Motion, the Plaintiffs shifted to a fire causation opinion that relies almost entirely on the presence of radiant flames to ignite the wooden stand. The Court's concern centers on the clear case law that rejects pyrolysis as an ignition source and the fact that the experts' test and conclusions appear to rely on this aspect of pyrolysis.
In Farmers Truck Insurance v. Magnetek, Incorporated, Justice Blackburn excluded expert testimony based on theories of pyrolysis because "the hypothesis for long term, low temperature ignition of wood cannot be considered to be a reliable basis for the admission of expert testimony . . ." In that case, the Plaintiffs sought to introduce expert testimony that relied on theories of pyrolysis to explain how heat from a fluorescent light ballast ignited furring strips in the ceiling. Applying the same Daubert reliability analysis Blackburn ruled that the "court's application of the standards of Rule 702 and the related cases is "a flexible and commonsense undertaking in which the trial judge is granted broad latitude in deciding both how to determine reliability as well as in the ultimate decision of whether the testimony is reliable." A trial judge must apply a "flexible" analysis to determining the reliability of an expert's testimony, and is not required to consider any one factor in particular, but instead to consider the relevant factors of each particular case.
C.A. No. 00-RB-2218 (CBS) (D. Col. 2002).
See id. at 2, quoting Smith v. Ingersoll-Rand Co., 214 F.3d 1235, 1243 (10th Cir. 2000).
See Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 152 (1999); M.G. Bancorporation v. Le Beau, 737 A.2d 513, 522 (Del. 1998).
Blackburn held that the record in Farmers did not produce "reliable principles and methods . . . for determining if and when wood exposed to the temperatures at issue in this case might catch on fire." The record in this case is equally undeveloped. The report and depositions indicate that pyrolysis was the experts' original theory of fire causation. Frederick explained that observing signs of pyrolysis provided the conclusion for his test. The experts' test was conducted and concluded on the basis of pyrolysis. They relied on pyrolysis as the ignition source of this fire. No general scientific principles have been presented to this Court supporting ignition by pyrolysis. Moreover, the experts' application and explanations of the role of pyrolysis in this fire are insufficient to consider their theory reliable.
See id. at 13.
(Frederick Depo. at 123-124).
The experts have failed to properly explain how pyrolysis contributed to this fire. First, they did not explain the time period and conditions that were necessary to reduce the ignition point of the wooden stand to approximately 193 degrees Fahrenheit, the highest temperature they recorded during their experiment. More importantly, they did not give a consistent and reliable basis for ignition once the ignition temperature of the wood had reached 193 degrees Fahrenheit. Frederick and Schneiders have never clearly and consistently explained their application of pyrolysis to the facts of this case. The failure of Schneiders and Frederick to explain the conclusions of their test indicates that whether or not the wooden stand underwent pyrolysis and was capable of ignition remains in doubt. Perhaps this explains why the Plaintiffs' primary basis for fire causation has now shifted to radiant flames. The Farmers decision casts a serious shadow on the experts' reliance on pyrolysis. But even if the experts' reliance on pyrolysis was proper, their total failure to apply the theory to the facts of this case destroys the reliability and relevance of their testimony.
The Motion is granted as to the exclusion of testimony by Frederick and Schneiders that the Rheem hot water heater started the fire at issue in this case due to a product defect. This decision is based on the Court's determination that the testing process, the reporting by the experts and the reliance upon pyrolysis as an explanation for fire causation independently and collectively produce an unreliable conclusion. For these reasons, the Defendant's Motion to Exclude the Expert Fire Causation Opinion at trial is GRANTED.