RULINGS GRANTING WHITE-RODGERS’ MOTION FOR SUMMARY JUDGMENT AND DENYING THER-MOGAS MOTION FOR SUMMARY JUDGMENT, AND ORDER
VIETOR, District Judge.
Plaintiffs,'John Stibbs and his immediate family, bring claims of negligence, breach of implied warranty, strict liability for placing a defective product in the stream of commerce, and fraud ágainst defendants based on damages caused by an explosion . that occurred when John Stibbs attempted to light his propane fueled water heater. Defendant White-Rodgers Division of Emerson Electric (“White-Rodgers”) moves for summary judgment, and Thermogas Company (“Thermogas”) joins in that motion. Plaintiffs resist, and oral hearings were held (at the same time as a similar motion by defendant Mid-America Pipeline Co.). The motion is submitted.
Summary Judgment Standard
Rule 56 of the Federal Rules of Civil Procedure provides that summary judgment
shall be rendered forthwith if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving parties are] entitled to a judgment as a matter of law.
Fed.R.Civ.P. 56(c). To preclude the entry of summary judgment, the nonmovants must make a sufficient showing on every essential element of their case for which they have the burden of proof at trial.
Celotex Corp. v. Catrett,
477 U.S. 317, 322-23, 106 S.Ct. 2548, 2552-53, 91 L.Ed.2d 265 (1986);
Continental Grain Co. v. Frank Seitzinger Storage, Inc.,
837 F.2d 836, 838 (8th Cir.1988). Rule 56(e) requires the nonmoving parties to go beyond the pleadings and by affidavits, or by the depositions, answers to interrogatories, and admissions on file, designate “specific facts showing that there is a genuine issue for trial.” Fed.R.Civ.P. 56(e);
Celotex,
477 U.S. at 324, 106 S.Ct. at 2553;
Johnson v. Schopf,
669 F.Supp. 291, 295 (D.Minn.1987). The quantum of proof that the nonmoving parties must produce is not precisely measurable, but it must be “enough evidence so that a reasonable jury could return a verdict.for the nonmovant[s].”
Anderson v. Liberty Lobby, Inc.,
477 U.S. 242, 257, 106 S.Ct. 2505, 2514, 91 L.Ed.2d 202 (1986);
Johnson,
669 F.Supp. at 295-96.
On a motion for summary judgment, the court views all the facts in the light most favorable to the nonmoving parties, and gives those parties the benefit of all reasonable inferences that can be drawn from the facts.
United States v. City of Columbia, Mo.,
914 F.2d 151, 153 (8th Cir.1990);
Woodsmith Publishing Co. v. Meredith Corp.,
904 F.2d 1244, 1247 (8th Cir.1990).
Facts
For purposes of this motion, the following facts are undisputed or represent plaintiffs’ version. John Stibbs was injured in a liquified petroleum (LP, referred to interchangeably by the parties and in this opinion as “propane”) explosion on July 17, 1992 while trying to light a water heater pilot light in. his basement. The LP system in the Stibbs’ home consisted of a 500 gallon LP tank, a RegO flow regulator, a kitchen stove, a furnace, and a State Industries water heater with a White-Rodgers control. During the five years Stibbs had owned the water heater, there had been no problems prior to the explosion.
The water heater and furnace were in one room of a small two-room basement accessible only through an exterior door. On July 15, 1992, the Stibbs ran out of hot water. John Stibbs went to the Thermogas office, stated that he was out of gas, and paid for 100 gallons of LP, which was delivered by Thermogas sometime on July 15 or 16. Thermogas contends that delivery was made on July 15. The delivery was made by Thermogas employee Larry Coy. Coy did not follow the proper “out of gas” procedure. Coy was not authorized to make delivery of the LP gas without supervision, as he had not completed the company’s training program. Plaintiffs’ expert, Tim Dunn, testified that had the propane leak been present at the time Thermogas delivered the propane, and if Thermogas delivered the gas on July 15, the amount of the damage from the explosion would have been greater.
At 10:30 a.m. on July 17, John Stibbs went to light the pilot on the water heater. As he descended the stairs to. the basement, he tested his lighter. The door between the two basement rooms was open five to six inches; he opened it the rest of the way and propped it open. He removed the water heater burner cover plate and crouched in front of the water heater. He did not smell gas. As he was looking at the lighting instructions, he lit his lighter and an explosion occurred.
White-Rodgers manufactured the water heater control on the Stibbs’ water heater in 1988. The control regulates the flow of gas to the pilot and the main burner. The control is designed to allow safe lighting of the pilot and to shut off gas flow if the pilot is extinguished. A user lights the pilot by turning the control knob to pilot and pushing it down. This manually opens the safety valve, allowing gas to flow to the pilot. Once the user lights the pilot, the button should be
held down for one minute. In this time, the flame heats a thermocouple. The heated thermocouple generates a current which in turn energizes a magnet to hold the safety valve open. Should the pilot go out, the thermocouple cools and the magnet de-energizes, allowing the safety valve to snap shut, blocking the gas flow. The control is also equipped with a fine-mesh filter and a deflection baffle to prevent debris in the gas stream from getting into the control mechanism. As part of its manufacturing process, White-Rodgers inspects and tests every control to ensure it works properly. This includes a leakage test on every control.
Plaintiffs’ expert, Dr. Jerry Hall, inspected the accident site on August 29, 1992. Hall examined the basement and tested the gas system, appliances, and furnace and water heater controls. Hall found that the gas pipes did not leak, and found no evidence that anyone had opened the lines. The water heater control did not leak during Hall’s on-site testing. Hall found the damage in the basement was consistent with a small accumulation of propane.
On May 7, 1993, a White-Rodgers representative disassembled the control in the presence of Hall and plaintiffs’ expert, Dr. Bruce Johnson. The control showed no signs of prior disassembly. Particulate matter and oxidation were present in the main body cavities of the control. There was a mark on the sealing surface of the safety valve which plaintiffs’ experts concluded was evidence of an indentation. The control did not leak, however, and the safety valve was operational. No particles were seen on the safety valve or the rubber valve seat.
Plaintiffs’ experts concluded that a particle from inside the control, which was most likely introduced at the time of manufacturing, had become lodged against the rubber seat of the safety valve either during the July 1992 repressurization or a prior refilling, allowing a small leak. The particle was then dislodged by the July 1992 repressurization or the explosion. Hall also testified at deposition, however, that he had no criticism of the design of the control, could not eliminate the possibility that the furnace control leaked, and could not eliminate the possibility that the alleged particle came into the control through the gas line. Hall did not test the force of the safety valve spring, the properties of the rubber valve seat, the size of a particle that could pass through the control valve mesh filter, or the air exchange rate of the basement. Expert Johnson admitted he does not know about White-Rodgers’ manufacturing process or quality control procedures, did not take any measurements of the alleged mark seen on the valve seat, did not test the force of the valve spring, and does not know the property of the valve seat or how it is manufactured. Johnson admitted a particle could cause a mark without causing a leak.
In March 1982, Emerson Electric made a presentation to the National LP Gas Association. In that presentation, Emerson acknowledged the need for change within the industry to provide more safety for the consumer. Emerson stated that all in the chain of distribution, including control manufacturers, are under a duty to educate and warn consumers as to the dangers involved in LP gas use. Emerson went on to cite the major causes of explosions in the use of LP gas, including,
inter alia,
the homeowner’s lack of knowledge of the dangers of LP gas and the inability of some people to recognize the smell of LP gas. Emerson had done nothing to warn end users of these problems.
Discussion
Although the issues are similar, I will address the White-Rodgers and the Thermogas motions separately, beginning with White-Rodgers.
White-Rodgers
Plaintiffs accept the White-Rodgers facts regarding the events surrounding the explosion and the general details regarding the White-Rodgers control. The alleged factual disputes are limited to the condition of the control, the evidence of causation, and the duty to warn or the adequacy of that warning. Plaintiffs allegations as to these issues are based on the testimony of their experts, and in particular Drs. Hall and Johnson. Defendants ask that I disregard that evidence as inadmissible under the analysis of
Daubert v. Merrell Dow Pharmaceu
ticals, Inc.,
509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).
Under
Dcmbert,
the trial judge evaluating expert testimony under Federal Rule of Evidence 702 must determine at the outset whether the expert is proposing to testify to:
(1) scientific knowledge that (2) will assist the trier of fact to understand or determine a fact in issue. This entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.
Id.
at 592, 113 S.Ct. at 2796. The Supreme Court provided to trial courts some general guidance for conducting this inquiry. First among the factors the Court considered relevant is whether a theory or technique can be (and has been) tested.
Id.
at 593, 113 S.Ct. at 2796.
As I read the various depositions of plaintiffs experts, and the parties’ treatment of those depositions, I find that a fair synthesis of the experts’ testimony is as follows. Plaintiffs’ experts hypothesize that a particle of unknown size and composition, most likely a small fragment of metal from the manufacture of the control, was inside the control. When the LP tank was refilled, the initial surge of gas was at a higher pressure than normal until the regulator adjusted itself to the-flow. This pressure moved the particle from its harmless resting place to the rubber seat of the safety valve. The particle kept the safety valve from sealing completely, allowing a very small amount of gas to leak into the basement. Because the basement was small and had no windows, there was little air movement and the gas concentrated near the control. When John Stibbs went to light the pilot light, this small quantity of gas exploded. The force of the explosion shook the particle loose from the rubber seat, and the particle flowed out of the control, leaving only a small mark on the rubber seat and allowing the control to return to normal functioning prior to post-explosion testing.
Plaintiffs argue that no “controlled laboratory tests” are necessary to support this theory of causation. Hall and Johnson have unchallenged credentials. They observed and investigated the site of the explosion and the control, they applied their background knowledge to what they saw and learned, and they engaged in sound reasoning to reach certain conclusions. Plaintiffs argue that this process was sufficient under Rule 702. I disagree. While there is no one right way to achieve scientific or technical validity in the analytical process, I conclude that the failure of Hall and Johnson to engage in testing of critical aspects of their hypothesis undermines the validity of their resulting “particle theory” of causation.
Hall and Johnson’s hypothesis is built on two objective facts: the presence of a mark on the rubber safety valve seat, and the absence of leaks in any other part of the LP system in the Stibbs home. With respect to the mark on the rubber safety'valve seat, the experts did not make any measurements of its size or depth. Hall Depo., May 22, 1995, at 28; Johnson Depo., Jan. 5, 1995, at 43. The experts did not determine the hardness or elasticity of the rubber used in the valve seat. Hall Depo., Oct. 11, 1995, at 27; Johnson Depo., Jan. 5, 1995, at 175. Indeed, there is no evidence that the mark represented an indentation. Hall testified that oxidation of aluminum can leave a fragile white powder. Hall Depo., Jan. 6, 1995, at 123-24. Hall and Johnson testified that it would be possible for a particle to leave a mark on the rubber without causing a leak because the rubber could effectively surround the material. Hall Depo., Oct. 11, 1995, at 176; Johnson Depo., Jan. 5,1995, at 206. Johnson also conceded that he was unfamiliar with the molding process for the rubber seat and did not know if the molding process resulted in any residual marks on the seat.
Id.
at 183.
Hall and Johnson theorize that the mark was caused when a particle, resting on the valve seat, was pressed against the rubber when the valve spring closed the safety valve. They admit that they do not know the composition of the particle or its hardness. Hall Depo., Jan. 6, 1995, at 121-124; Johnson Depo., Jan. 5,1995, at 133. They did not test the force of the White-Rodgers control valve spring. Hall Depo., Jan. 6, 1995, at 122; Johnson Depo., Jan. 5, 1995, at 44-45. Hall did not test to see the effect or marks made
by the impact of various particles on the rubber disk. Hall Depo., May 22, 1995, at 26. Of particular significance to White-Rodgers’ liability, the experts cannot state with certainty the source of the alleged particle. It may have come from the manufacturing process, from the gas supply (if it was small enough to get through the filter screen), or from the control being taken apart or being flooded. Hall Depo., Jan. 6, 1995, at 117-20; Johnson testified that he was unfamiliar with steps White-Rodgers takes in the manufacturing and quality control processes to eliminate residual particles in the controls. Johnson Depo., Jan. 5,1996, at 60.
Hall and Johnson suggest that either the particle or the indentation left by the particle prevented the safety valve from sealing properly, allowing a slow leak of propane into the basement, where it achieved sufficient concentration to explode. They have not tested or calculated the rate of air exchange in the basement. Hall Depo., May 22, 1995, at 45-47; Hall Depo., Jan. 6,1995, at 126; Johnson Depo., Jan. 5,1995, at 186. The particle that allegedly caused this leak was not found in post-accident investigation of the valve. Hall and Johnson state two theories for why the particle is no longer present: it was pushed away by the pressure of refilling the tank, or it was shaken loose in . the explosion. Hall has not conducted any tests to attempt to simulate either condition. Hall Depo., Jan. 6, 1995, at 128. Johnson admits he does not know what became of the particle, and adds the possibility that a fireman could have pressed the reset on the gas control and dislodged the particle. Johnson Depo., Jan. 5,1995, at 12.
Hall and Johnson surmise that because the LP system passed leak tests after the explosion, the leak must have come from a control which had since returned to operating condition. Both experts concede, however, that they cannot rule out the furnace control. Hall Depo., Oct. 11, 1995, at 134; Johnson Depo., Jan. 5,1995, at 8. It is also undisputed that the White-Rodgers water heater control has never leaked in post-explosion tests. Hall Depo., Jan. 6,1995, at 53-54.
Plaintiffs argue that “differential diagnosis”—determining causation by eliminating possible causes until only one remains—is a legitimate scientific method, one often used in medical analysis, and is sufficient “testing” to support the conclusions of Hall and Johnson. The Eighth Circuit has previously addressed this argument.
That inference turns scientific analysis on its head. Instead of reasoning from known facts to reach a conclusion, the experts here reasoned from an end result in order to hypothesize what needed to be known but was not.
While it may be that this sort of reasoning could pass muster in some cases where the obvious result explains the etiology * * * such reasoning cannot apply here where several possible causes could have produced one effect.
Sorensen v. Shaklee Corp.,
31 F.3d 638, 649 (8th Cir.1994). The “differential diagnosis” is particularly problematic in this case because in actual post-explosion testing of the White-Rodgers control it worked correctly.
More importantly, even if the broad outlines of the experts’ procedure are acceptable in this case, the implementation of the investigation is not.. Rule 702 permits admission of scientific or technical knowledge. “[T]he word ‘knowledge’ connotes more than subjective belief or unsupported speculation.”
Daubert,
509 U.S. at 590, 113 S.Ct. at 2795. The testimony here lacks the support to rise above speculation because it remains untested in many important respects. Basic matters such as whether the molding process of the rubber valve seat could leave a mark, whether there was an actual indentation or just a surface mark, what properties the rubber used by White-Rodgers possessed, and what force the safety valve spring generates were all left unanswered, as were many other relevant questions.
It is important, under
Daubert,
to consider whether the premises on which experts base their conclusions have been tested for validity. In this ease, they were not. It is also important to consider whether an expert’s processes and conclusions can be verified by subsequent, independent testing for “falsifiability, or refutability * *
Daubert,
509
U.S. at- 593, 113 S.Ct. at 2796 (citing K. Popper,
Conjectures and Refutations: The Growth of Scientific Knowledge,
37 (5th ed. 1989)). The testimony of Hall and Johnson is also lacking in this respect, a point Johnson concedes.
Q. [by David Shinkle, attorney for defendant Natural Gas Odorizing, Inc.].
Well, is there any way to disprove your theory here in light of the fact that the valve doesn’t leak now and the safety properly seats? Is there any way to disprove that? What you are saying is * * * the safety failed to seat and there was a particulate matter in there and then the explosion occurred. And it took the particulate matter out and now the safety works properly again. Would you agree with me that * * * effectively prevents anybody from disputing your theory?
A. * * * [Y]eah, it makes it tough to dispute it. It’s very hard to prove a negative, I understand that.
Johnson Depo., Jan. 5, 1995, at 135. Such “essentially irrefutable” opinion evidence is suspect because it cannot be verified.
See Gier v. Educational Serv. Unit No. 16,
845 F.Supp. 1342, 1348 (D.Neb.1994) (excluding plaintiffs’ expert testimony, stating, “In such a ease, the expert’s conclusions are as impenetrable as they are unverifiable.”),
aff’d,
66 F.3d 940 (8th Cir.1995).
I conclude that the testimony by Drs. Hall and Johnson relied on by plaintiffs is inadmissible under Rule 702 and
Daubert
for the purpose of showing that a particle in the White-Rodgers control caused a propane leak and subsequently and explosion. Absent that evidence, plaintiffs have not met their burden to make a sufficient showing on every essential element of their case for which they have the burden of proof at trial.
Celotex,
477 U.S. at 322-23, 106 S.Ct. at 2552-53. There remains no evidence from which a jury could find or infer that the White-Rodgers control was a proximate cause of plaintiffs’ injuries. This conclusion precludes each of plaintiffs’ theories of liability: negligence in manufacture or in failing to warn, strict product liability, and breach of implied warranty. Accordingly, defendant White-Rodgers’ motion for summary judgment will be granted.
Thermogas
Plaintiffs allege that Thermogas (referred to as MAPCO Natural in the Complaint) was negligent in failing to follow its own procedures or take other reasonable safety measures in the delivery of LP to Stibbs. Plaintiffs also assert that Thermogas breached an implied warranty by delivering unsafe LP (based on plaintiffs’ argument that the LP was improperly odorized), and that it is strictly liable for placing in the stream of commerce a defective product.
Thermogas argues that if there is no valid expert testimony as to causation, proximate cause fails as to Thermogas as well. Because Hall and Johnson can only speculate as to the cause, rate, and time of origin of the leak, there is no evidence that the Thermogas employee could have, under any circumstances, discovered the leak at the time of delivery. Indeed, Thermogas argues that absent Hall and Johnson’s testimony, there is no evidence that propane was the cause of the explosion. Plaintiffs argue that the fact of an explosion in the basement while John Stibbs was trying to light a propane fueled appliance raises a genuine issue of fact as to the presence of a gas leak. Plaintiffs further argue that the evidence that Thermogas employee Larry Coy, who delivered the propane to the Stibbs home, did not follow procedures for an “out-of-gas” situation or for a new, customer raises a genuine issue of fact as to whether Thermogas acted negligently in a way that resulted in (or failed to prevent) plaintiffs’ injuries. .
As an initial matter, I conclude that there is sufficient evidence to permit a finding that there was a propane explosion. The circumstances surrounding the explosion sufficiently suggest the cause was propane. Hall and Johnson also have testified that, in then-opinion, the explosion was a propane explosion. While I have concluded above that the testimony of Hall and Johnson is not-admissible as to their theory of causation,
Daubert
does not require me to take an “all or nothing” approach to the testimony of experts. I find plaintiffs’ experts credible on the issue of whether' the explosion was an LP gas
explosion, and the circumstances of the explosion.
Thermogas argues that, as with White-Rodgers, there is no evidence establishing Thermogas’ actions as a proximate cause of plaintiffs’ damages and therefore all of plaintiffs’ claims against Thermogas fail. Plaintiffs argue that Coy’s failure to follow proper procedures resulted in a failure to detect or resolve an unsafe condition in the LP system. Thermogas suggests that this theory requires some evidence that the condition would have been discoverable if Coy had properly checked. It is an undisputed fact that there had been no prior problems with the LP system. It is also undisputed that post-explosion testing revealed no leaks or detectable flaws in the system.
Plaintiffs advance other arguments, however, that do not rely on Coy’s ability to detect a leak at time of delivery. Plaintiffs present evidence that it is Thermogas’s policy when refilling an “out-of-gas” system to relight the pilot lights for the customer. If the customer is not home, the supply valve from the LP tank to the home should be closed and a note left warning the customer not to attempt to light the pilots, but rather to call Thermogas to come back and restart the appliances.
See
Affidavit of Robert F. Leonard Filed With Plaintiffs’ Resistance to Thermogas Company’s Motion, Exhibit B. Plaintiffs assert that had a professional come back to relight the appliances, that person would have had a better chance of detecting the leak—whatever its origin—or ensuring that the appliance controls were operating properly, particularly if the main supply valve had remained closed until that time. Larry Coy does not remember shutting off the Stibbs’ supply valve. Coy Depo., Jan. 4, 1995, at 30. It also appears that Coy did not take any particular safety steps, in part because he claims to have been unaware that the Stibbs were completely out of gas. Coy Depo., Jan. 4, 1994 at 21, 30, 47. I conclude that there remain genuine issues of material fact as to Thermogas’ negligence in the refilling of Stibbs’ LP tank and the proximate cause of such negligence in bringing about the explosion and resulting damages. Accordingly, Thermogas’ motion for summary judgment will be denied as to plaintiffs’ negligence claims.
Thermogas does not separately address plaintiffs’ implied warranty or strict liability claims. While there may be arguments other than lack of proximate cause that could apply to those claims, they were not raised by Thermogas in its joinder with White-Rodgers. Accordingly, Thermogas’ motion for summary judgment will be denied as to those claims as well.
Rulings and Order
Defendant White-Rodgers’ motion for summary judgment is GRANTED. Defendant Thermogas’ motion for summary judgment is DENIED.
IT IS ORDERED that plaintiffs’ complaint is DISMISSED as to defendant White-Rodgers.