POPE, Chief Justice.
This is a products liability case in which the plaintiffs William C. Kindred and Ernest J. Kurtz pleaded strict liability on both a design defect theory and a marketing defect theory. The trial judge submitted special issues only on the marketing defect. After the jury answered the special issues on marketing defect in the negative, the trial court rendered a take nothing judgment. The court of appeals affirmed. 644 S.W.2d 828 (Tex.App. — Corpus Christi 1982). We reverse the judgment of the court of appeals and remand the cause to the trial court for a new trial.
A representative of Con/Chem, Inc. visited Spoetzl Brewery, Inc. in Shiner, Texas, and recommended the use of his company's paint for the inside of the beer storage tanks. Spoetzl had problems with the paint because it would not adhere. The Con/Chem representative had Spoetzl send paint samples. Con/Chem then recommended that Spoetzl use their primer, Cono/Prime X, prior to applying the paint. Cono/Prime X is the product alleged to be defective.
Plaintiffs Kindred and Kurtz were employed by Spoetzl as maintenance workers. On April 28, 1977 the two men were working inside one of Spoetzl's cylindrical storage tanks. The only openings in the tank were a two-inch valve at the top and a "manhole" at the bottom.
The two men sandblasted, vacuumed and wiped the inside surface of the tank and had been applying the primer for approximately 20 minutes when a fire started. The men had to pass through the fire to exit. They sustained serious burns to their arms, legs, hands and faces.
The record showed that Cono/Prime X has a flash point of 53 degrees Fahrenheit, which means that in the presence of oxygen and an ignition source the material would burn when warmed to approximately 53 degrees. The evidence showed that the temperature in the room containing the tank was about 48 degrees and that the temperature inside the tanks was about 50 degrees. None of the witnesses could identify the exact ignition source.
Kindred and Kurtz alleged (1) a marketing defect involving the failure to adequately warn of the dangers of using the product and (2) a design defect rendering the product unreasonably dangerous. They requested special issues on both alleged defects, and they tendered issues to the trial court in substantially correct form. However, the trial court allowed the submission of special issues only on the failure to warn. The jury answered those issues in favor of Con/Chem. The trial court then rendered a take nothing judgment against Kindred and Kurtz. The court of appeals affirmed, holding that Kurtz and Kindred failed to present more than a scintilla of evidence on defective design, and therefore the trial court properly refused to submit the issues on design defect. Kindred and Kurtz present a single point of error in which they claim that they presented some evidence to support their requested special issues on design defect. We agree.
In Turner v. General Motors Corp., 584 S.W.2d 844, 847 (Tex. 1979) we stated a test for determining design defect. That test requires balancing the utility of the product against the risks involved in its use in order to find that the design is unreasonably dangerous. The next year in Boatland of Houston, Inc. v. Bailey, 609 S.W.2d 743 (Tex. 1980), we stated,
The jury may consider many factors before deciding whether a product's usefulness or desirability are outweighed by its risks. Their finding on defectiveness may be influenced by evidence of a safer design that would have prevented the injury. [Citations omitted.] Because defectiveness of the product in question is determined in relation to safer alternatives, the fact that its risks could be diminished easily or cheaply may greatly influence the outcome of the case.
. . . This feasibility is a relative, not an absolute, concept; the more scientifically and economically feasible the alternative was, the more likely that a jury may find that the product was defectively designed. A plaintiff may advance the argument that a safer alternative was feasible with evidence that it was in actual use or was available at the time of manufacture. Feasibility may also be shown with evidence of the scientific and economic capacity to develop the safer alternative.
Id. at 746.
In attempting to meet this test, Kindred and Kurtz offered the expert testimony of Dr. Key, an environmental chemist. Dr. Key testified that using Cono/Prime X inside the storage tank, as Con/Chem had recommended, created a "very dangerous" situation. Dr. Key also testified that Cono/Prime X consisted of five percent ethanol, ninety percent isopropyl alcohol, and five percent silane. He explained that the ethanol and alcohol, ninety-five percent of the product, constituted the "carrier" which gave the needed fluidity for application and which evaporated in a process called polymerization. Dr. Key testified that the "carrier" mixture of ethanol and alcohol was sufficiently flammable to cause the product, in his opinion, to be defective. Dr. Key further testified that water, a nonflammable substance, could have been used instead of the ethanol and isopropyl alcohol, both of which were "considerably" more expensive than water. Dr. Key pointed out that if water was used, a more expensive silane, five percent of the product, would be needed. Additionally, Dr. Key testified that the use of water would not impair the product's usefulness.
Con/Chem argues that this testimony was no evidence that Cono/Prime X, as designed, was unreasonably dangerous, but rather, at best, it was some evidence that Cono/Prime X was unreasonably dangerous as a result of failure to warn. We disagree. Dr. Key's testimony relates to the feasibility of a safer alternative and is some evidence of a design defect.
Con/Chem also argues that Kindred's and Kurtz's reasoning assumes that the cost of ingredients was the only variable in the change to a safer alternative and that there was no evidence of the cost of the final product of the safer alternative. Kurtz and Kindred do not have to prove the final cost of the safer alternative to get to the jury. They are only required to present some evidence of probative value. Expert testimony that water could have been substituted for ninety-five percent of the product is evidence of probative value. Con/Chem was silent as to any expense in producing the safer alternative. The jury therefore could have inferred that Cono/Prime X could have been made less flammable at a reasonable cost.
Con/Chem also says that there was no evidence that Cono/Prime X was the producing cause of the fire. Although no witness could conclusively state that the product was the cause, the jury could have made that determination from circumstantial evidence. General Motors v. Hopkins, 548 S.W.2d 344 (Tex. 1977); Birmingham v. Gulf Oil Corp., 516 S.W.2d 914 (Tex. 1974); Darryl v. Ford Motor Co., 440 S.W.2d 630 (Tex. 1969); Mueller Co. v. Corley, 570 S.W.2d 140 (Tex.Civ.App. — Houston [1st Dist.] 1978, writ ref'd n.r.e.). The testimony indicates that Cono/Prime X, as designed, was a "very flammable mixture," which was being applied inside the tank, as Con/Chem intended, when the fire occurred.
We hold that the evidence was sufficient to submit the issue to the jury. Tex.R.Civ.P. 279. "The factual insufficiency of the evidence to support an affirmative answer to an opponent's issue furnishes no basis for refusal to submit the issue." Clarostat Manufacturing Inc. v. Alcor, 544 S.W.2d 788, 791 (Tex.Civ.App. — San Antonio 1976, writ ref'd n.r.e.). The trial court may decline to submit a relevant issue, only if there is no evidence to support it. Garza v. Alviar, 395 S.W.2d 821, 824 (Tex. 1965). When the evidence offered to prove a vital fact is so weak as to do no more than create a mere surmise or suspicion of its existence, the evidence is no more than a scintilla and, in legal effect, is no evidence. Seideneck v. Cal Bayreuther Associates, 451 S.W.2d 752, 755 (Tex. 1970); Joske v. Irvin, 91 Tex. 574, 581-82, 44 S.W. 1059, 1062 (1898). See Calvert, "No Evidence" and "Insufficient Evidence" Points of Error, 38 Tex.L.Rev. 361 (1960) and 5 W. Dorsaneo, Texas Litigation Guide, § 122.03(2)(a) (1983). The test for the application of this no evidence/scintilla rule is that if reasonable minds cannot differ from the conclusion that the evidence offered to support the existence of a vital fact lacks probative force, it will be held to be the legal equivalent of no evidence. However, there is some evidence, more than a scintilla, if the evidence furnishes some reasonable basis for differing conclusions by reasonable minds as to the existence of the vital fact. Martinez v. Delta Brands, Inc., 515 S.W.2d 263, 265 (Tex. 1974). See, Henderson v. Ford Motor Company, 519 S.W.2d 87, 95-96 (Tex. 1974) (Johnson, J. dissenting).
We hold that Dr. Key's testimony was more than a scintilla because it furnished a reasonable basis for differing conclusions concerning the existence of a design defect in Cono/Prime X. The special issues on design defect should have been submitted to the jury.
The judgments of the court of appeals and the trial court are reversed, and the cause is remanded to the trial court for a new trial.