No. 76-233.

87 Wis.2d 243 (1979)

274 N.W.2d 647

COGSWELL, and others, Plaintiffs, v. ROBERTSHAW CONTROLS COMPANY, and another, Appellants: ITTGENERAL CONTROLS, INC., Respondent.

Supreme Court of Wisconsin.

Decided January 30, 1979.

Attorney(s) appearing for the Case

For the appellants the cause was submitted on the brief of Doar, Drill, Norman, Bakke, Bell & Skow of New Richmond.

For the respondent the cause was submitted on the brief of Wilcox & Wilcox of Eau Claire.


This is an appeal from a judgment dismissing the third-party complaint and cross-complaints of the defendants Robertshaw Controls Company and The Travelers Insurance Company (appellants) against defendant ITT-General Controls, Inc. (respondent), entered July 20, 1976, the Honorable Robert F. Pfiffner, presiding.

The question on appeal is: Was the finding by the trial judge that a defect in the Unitrol valve in the water heater manufactured by Robertshaw caused an explosion injuring the plaintiffs against the great weight and clear preponderance of the evidence?

This case involves a dispute between two manufacturers as to whose product was responsible for causing a gas explosion in the basement of a summer cabin on Upper Eau Claire Lake in Bayfield County.

In September of 1965, Colonel Roger K. Peterson and Lillian Peterson acquired the cabin which they used over a period of five years as a vacation home. In the basement of the cabin there was a propane gas fired furnace containing an ITT-General Control B52 valve and pilot relay valve. There was also a water heater in the basement containing a Robertshaw Unitrol 110 control which regulated the flow of gas to the main burner and the pilot burner. In the case of each appliance, the valve was designed to shut down the flow of gas in the event that the pilot light was extinguished. In addition, there was an electrically activated water pump in the basement.

Colonel Peterson followed a routine of shutting down the gas and water system at the end of the vacation season every fall, and re-establishing the systems in the following spring. The first time he lit the water heater after acquiring the cabin, he had trouble lighting the pilot light, but thereafter he never had any trouble with either appliance until the accident. Colonel Peterson was the only one who either activated or shut down the furnace or water heater during the period from 1965 to 1971, except for one occasion when he contracted with a local plumbing and heating company to start up the systems.

At Easter, 1971, Colonel Peterson re-established the flow of gas and water into the house and lit the burners in the furnace and water heater. From that time on until the accident, July 20, 1971, no one manipulated the controls on the water heater or the furnace.

A few days before the accident, Mrs. Peterson and her two children went up to the cabin. Colonel Peterson was away on assignment with the Air Force. On the evening of July 20, 1971 around 9:00 or 9:15 p.m. Mrs. Peterson ran water in the kitchen sink to clean up the kitchen when she noticed that there was no hot water. On inquiry, she learned that her daughter had run a bath at about 7:00 p.m. and the water "wasn't all that warm." Mrs. Peterson went down to the basement to investigate and saw that there was no flame in the pilot or main burner of the water heater. However, she saw that the dial on the control of the water heater was pointed to the "On" position. She attempted to turn the dial to "Off," but she could not move it. She did not check to see whether the pilot light in the furnace was on. The weather was cool enough to wear jackets at night.

Mrs. Peterson and her son Roger Leland Peterson left the cabin to get assistance from a neighbor. The neighbor, Mr. Merton Cogswell, came back to the Peterson cabin, along with Mr. Harold Robbins. Mrs. Peterson, her son Roger, Mr. Cogswell and Mr. Robbins went down to the basement of the cabin. Mrs. Peterson testified that the furnace was not running at the time they entered the basement. No one testified to smelling gas upon entering the basement. Mr. Robbins testified that he smelled gas for the first time at the instant of the explosion.

Mr. Cogswell went over to the water heater and examined the control. The dial was pointed toward "On." He saw no flame in the pilot or main burner. He neither heard nor smelled gas escaping from the main burner or the pilot. He could not turn the gas cock dial to the "Off" position. He next depressed the red reset button, and with the button depressed, he succeeded in turning the dial first counterclockwise, and then clockwise to the "Off" position. He continued to depress the red button, and knelt down with his face within six inches of the main burner. He neither heard nor smelled gas escaping. With his finger on the red button, he again turned the dial to "On." As he turned the dial from "Off" to "On," he heard a rumble, followed immediately by an explosion. He said that the rumble sounded like the ignition of gas, like a gas fired furnace going on. Mr. Cogswell variously estimated the amount of time he spent in the basement as two to four minutes on one occasion and six minutes on another occasion. He testified that he did not hear the water pump switch on when he was in the basement. Mr. Robbins testified that he heard a swish and a click almost simultaneous with the explosion. He said that he was familiar with water pumps and the sound he heard was not one he would associate with a water pump running. He smelled gas at the time he heard the swish and click.

Kenneth Atherholt, assistant to the general manager of Robertshaw Controls, explained that the Unitrol 110 valve on the water heater was designed to interrupt the flow of gas from the storage source through the control in the event that the pilot light were to go out. The valve is located immediately below the red button on the top of the control. When the red button is depressed manually, it opens the valve. If the dial is in an "Off" position, gas from the storage source will flow only as far as the gas cock and no further. If the dial is at the "Pilot" position, the gas will flow from the storage source into the valve and down to the pilot. However, if the dial is in the "On" position, and the valve is open, the gas would flow from the storage source into the valve and to both the pilot and main burners. Mr. Atherholt said that under industry standards, the maximum permissible time between the extinguishing of the pilot burner flame and the close of the valve below the red button is three minutes. This time period is called dropout time. He testified that when the Unitrol 110 valve was tested subsequent to the accident, the valve shut down within a minute.

The ITT-General Control valve on the furnace was also tested for dropout time. Mr. Atherholt testified that four times the pilot light was lit and then extinguished, and it was still possible to relight the pilot without resetting the valve. From this he concluded the general valve was not functioning properly because it should not be possible to re-establish the pilot light once the thermocouple near the pilot light has cooled. The tests on the general valve were repeated at a later date with the same result, namely after the automatic pilot device should have dropped out there was still gas available to the pilot. However, the main burner was not relightable.

Atherholt testified that the Unitrol valve was designed with a safety interlock device which would prevent the operator from being able to depress the red button when the valve was in the "On" position. The reason for this design feature was to keep gas from flowing to the main burner when the pilot light is out. If the interlock device is functioning properly, the user should not be able to turn the gas cock dial from the "Pilot" position to the "On" position while the red button is depressed.

However, inspection showed that the safety interlock device was very rusty and scratched. The manufacturer had received complaints from consumers that they were able to depress the red button while the valve was in the "On" position. As a result of these problems, the company made some design changes in the Unitrol 110 valve in 1957, changing the thickness of one of the parts, an index plate, from twenty-four thousandths inch to forty thousandths inch and strengthening the reset button. The company made no effort to recall the older model valves. The valve in the Peterson cabin was of the older model, with the plate measuring from twenty-six to twenty-nine thousandths inches. There was no evidence that the valve had been forced with tools.

There was a conflict in the expert testimony as to how much gas was necessary to produce the explosion. The expert for Robertshaw Controls, Mr. Marvin Salzenstein, said that in his opinion from five to fifteen (more likely fifteen) cubic feet of gas would have had to be present in the basement in order to produce the explosion. Assuming rates of gas flow of 1.2 to 1.77 cubic feet per hour, he testified that the General Control pilot valve on the furnace would have had to be leaking from eight and one-half to twelve and one-half hours to produce fifteen cubic feet of gas in the basement.

Fulton Holtby, expert witness for ITT-General Controls, testified that enough gas could have leaked from the pilot and main burner of the water heater in two minutes to produce the explosion that occurred.

The trial court concluded that while both the Unitrol and the ITT-General valves were defective, it was the Unitrol valve in the water heater which caused the explosion. Further facts will be presented in the opinion.


Findings of fact by the trial court will not be upset on appeal unless they are against the great weight and clear preponderance of the evidence. The evidence supporting the findings of the trial court need not in itself constitute the great weight or clear preponderance of the evidence; nor is reversal required if there is evidence to support a contrary finding. Rather, to command a reversal, such evidence in support of a contrary finding must itself constitute the great weight and clear preponderance of the evidence. In re Estate of Jones, 74 Wis.2d 607, 611, 247 N.W.2d 168 (1976). In addition, when the trial judge acts as the finder of fact, and where there is conflicting testimony, the trial judge is the ultimate arbiter of the credibility of the witnesses. Gehr v. Sheboygan, 81 Wis.2d 117, 122, 260 N.W.2d 30 (1977). When more than one reasonable inference can be drawn from the credible evidence, the reviewing court must accept the inference drawn by the trier of fact. Id.

Two things were necessary to produce the explosion which took place in the Peterson cabin—a source of fuel and a source of ignition. Three factual possibilities were considered by the trial court:

"1. Both the pilot on the hot water heater and the furnace went out at the same time, or were out at the time of explosion, and both faulty valves permitted gas to escape into the basement (Unitrol 110 by manipulation of Cogswell, and perhaps at the time the pilot went out) creating an explosive mixture which was ignited by a spark from the motor on the water pump.

"2. Both pilots went out with the Unitrol 110 valve closing and the ITT-General pilot relay not closing with the resulting emission of gas from the ITT valve creating the explosive mixture which was ignited by a spark from the motor of the water pump.

"3. That the pilot burner on the furnace remained lighted at all times, and that the ignition of the main burner of the furnace in turn ignited and exploded gas that had escaped into the basement from the Unitrol 110."

[4, 5]

The trial court chose the third alternative. There is credible evidence in the record to support the choice of the trial court.

In the first place, no one smelled gas upon entering the basement. The theory advanced by Robertshaw Controls was that the valve in the furnace had been leaking for hours. The actual size of the pilot orifice was determined to be .011 inches, permitting a flow rate of only .38 cubic feet per hour—slower than the 1.2 to 1.77 cubic feet per hour assumed by the Robertshaw expert. At that rate, the expert conceded the General Control would have had to be leaking for thirty-nine hours before the explosion. Nobody ever looked to see whether the pilot light to the furnace was burning. The defect in the safety device in the ITT-General value would only be pertinent if, in fact, the pilot light was out. The record supports a conclusion that the pilot light in the water heater was more likely to have gone out than the pilot in the furnace. There was no protective cover over the hot water heater pilot, while the furnace pilot light was covered by a louvered door. Further, two witnesses, Mr. Cogswell and Mr. Robbins, testified that they heard what sounded like a gas furnace "kicking" on. Neither heard what sounded like a water pump switch on. The trial court commented in its opinion:

"Everyone in his daily experiences in life has been exposed to sounds which in turn are converted into meaning and cause to the listener. A gas appliance burner igniting has a distinctive sound. Both of these men described the sound as that of a gas furnace going on. I believe that both of them did hear the furnace go on. Their testimony is credible. Thus, I concluded or found that at the time of the explosion the main burner of the furnace lighted, and in order to do so the pilot on the furnace indeed had to be burning and had not gone out."

In addition, the temperature outside was between fifty and sixty degrees, cool enough to wear jackets. This is consistent with the conclusion that the furnace would be called upon for heat and would automatically turn on.

Robertshaw Controls argues that Mr. Cogswell was an incredible witness because he said that he turned the gas cock dial ninety degrees counterclockwise from the "On" position, something which was impossible, according to the testimony of Mr. Salzenstein. However, there was evidence that the safety interlock mechanism below the valve was badly rusted, and that the index plate, another component of the interlock, had been scratched when a slide mechanism had slipped over or under the plate.

Robertshaw argues that Mr. Cogswell was not at the controls of the water heater long enough for sufficient gas to leak out to cause the explosion. Cogswell had estimated that after he had depressed the red button, about a minute and a half passed before the explosion. He also estimated that of that one and one-half minutes with the red button down, the gas cock dial was pointed toward "On" for twenty-five to thirty-five seconds. However, the trial court pointed out that these attempts to fix the amount of time were merely estimates. For the period of time that the red button was depressed with the control in the "On" position, a full flow of gas was delivered to both the main burner and the pilot burner. The fact that nobody in the basement smelled gas until just before the explosion is more consistent with the theory selected by the trial court than the theory advanced by Robertshaw that the furnace control valve had been leaking for hours before the explosion. The fact that Mr. Cogswell was able to turn the dial to "On" while the red button was depressed demonstrated there was something wrong with the valve.


We conclude that the finding by the trial court that the defective condition of the Unitrol valve caused the explosion was not against the great weight and clear preponderance of the evidence.

By the Court.—Judgment affirmed.


1000 Characters Remaining reserves the right to edit or remove comments but is under no obligation to do so, or to explain individual moderation decisions.

User Comments

Listed below are the cases that are cited in this Featured Case. Click the citation to see the full text of the cited case. Citations are also linked in the body of the Featured Case.

Cited Cases

  • No Cases Found

Listed below are those cases in which this Featured Case is cited. Click on the case name to see the full text of the citing case.

Citing Cases