OPINION

Opinion by Justice MOSELEY.

Shaun T. Mian Corp. d/b/a Midway Tower, Irving L. Humphrey, and Calvin Otte, appellants, sued appellee Hewlett-Packard Company (“HP”), claiming a printer/fax machine it manufactured was defective and caused a fire that damaged their property. They also alleged HP was negligent in the design, manufacturing, and marketing of the machine. HP filed a combined “no-evidence” and traditional motion for summary judgment, see TexR. Civ. P. 166a(i), 166a(c), which the trial court granted. As discussed herein, we conclude appellants’ circumstantial evidence was sufficient to raise an issue of material fact as to each contested element of their manufacturing defect claim. Thus, we reverse the trial court’s judgment as to that claim and remand it to the trial court for further proceedings. We affirm the trial court’s summary judgment as to appellants’ other claims.

I. Standard of Review

A party may “move for summary judgment on the ground that there is no evidence of one or more essential elements of a claim or defense on which an adverse party would have the burden of proof at trial.” Tex.R. Civ. P. 166a(i). Assuming such a motion otherwise complies with the rule, it must be granted unless the non-movant produces summary judgment evidence raising a genuine issue of material fact as to the contested element or elements. Id.; see W. Invs., Inc. v. Urena, 162 S.W.3d 547, 550 (Tex.2005).

A party may also move for summary judgment on traditional grounds, i.e. there is no genuine issue as to a specified material fact and, therefore, the moving party is entitled to judgment as a matter of law. Tex.R. Civ. P. 166a(c). For a defendant to prevail on a traditional motion for summary judgment, it must either disprove at least one element of the plaintiffs claim as a matter of law or conclusively establish all elements of an affirma *855 tive defense to that claim. Friendswood Dev. Co. v. McDade + Co., 926 S.W.2d 280, 282 (Tex.1996). If the movant meets its burden, then and only then must the non-movant respond and present evidence raising an issue as to the material fact(s) in question. See Rhone-Poulenc, Inc. v. Steel, 997 S.W.2d 217, 222-23 (Tex.1999); Tex.R. Crv. P. 166a(c).

We review de novo a summary judgment granted on either no-evidence or traditional grounds, examining “the entire record in the light most favorable to the nonmovant, indulging every reasonable in-fei’ence and resolving any doubts against the motion.” Sudan v. Sudan, 199 S.W.3d 291, 292 (Tex.2006) (per curiam) (quoting City of Keller v. Wilson, 168 S.W.3d 802, 823 (Tex.2005)). We cannot affirm a summary judgment on grounds other than those specified in the motion. Tex.R. Civ. P. 166a(e), 166a(i). However, if the trial court’s order does not specify the grounds on which it granted summary judgment, we affirm if any of the grounds specified in the motion are meritorious. See Provident Life & Acc. Ins. Co. v. Knott, 128 S.W.3d 211, 216 (Tex.2003). When the motion for summary judgment presents both no-evidence and traditional grounds, we first review the propriety of the summary judgment under the rule 166a(i) no-evidence standards. See Ford Motor Co. v. Ridgway, 135 S.W.3d 598, 600 (Tex.2004) (appellate court reviews no-evidence summary judgment before addressing traditional summary judgment).

We affirm a no-evidence summary judgment if, as to an essential element of the claim or defense identified in the motion: (a) there is a complete absence of evidence; (b) the court is barred by rules of law or of evidence from giving weight to the only evidence offered; (c) the evidence offered is no more than a mere scintilla; or (d) the evidence conclusively establishes the opposite. See King Ranch, Inc. v. Chapman, 118 S.W.3d 742, 751 (Tex.2003) (citing Merrell Dow Pharms., Inc. v. Havner, 953 S.W.2d 706, 711 (Tex.1997)). 1 Thus, to avoid a no-evidence summary judgment, the non-moving party must bring forth more than a scintilla of probative evidence to raise a genuine issue of material fact as to the element or elements attacked. Id. More than a scintilla of evidence exists when the evidence “rises to a level that would enable reasonable and fair-minded people to differ in their conclusions.” Id. (quoting Merrell Dow Pharms., Inc., 953 S.W.2d at 711). On the other hand, the evidence amounts to no more than a scintilla if it is “so weak as to do no more than create a mere surmise or suspicion” of a fact. Id.

We affirm a traditional summary judgment if the evidence submitted in support of the motion and any response shows that no genuine issue of material fact exists, and the moving party is entitled to judgment as a matter of law. Tex.R. Civ. P. 166a(c).

II. Factual and Procedural Background

A. Background Facts

About two months before the fire, Otte purchased a new HP printer/fax machine from a retail merchant in Los Angeles. The printer arrived in a sealed box and was undamaged when delivered to Otte. Following the instructions provided with the printer, Otte unpacked and set up the printer in his office. He placed it on the left end of the credenza behind his desk, near the box housing his computer’s central processing unit (CPU). He plugged the printer into a power strip and placed *856 the printer’s power supply cord on the floor beneath the printer. The printer performed without apparent problem until the fire. Nothing was ever spilled on the printer, and it was never repaired, serviced, or modified. According to Otte’s affidavit, the printer was the only electrical device on the credenza that was plugged in, and the printer power supply was the only electrical device at floor level behind, beneath or beside the credenza near the printer.

About two months after Otte purchased the printer, a fire broke out in Otte’s office. The fire occurred on a Saturday morning, and no one was in the office at the time. The fire damaged Otte’s and other offices in the building. The printer was severely damaged in the fire.

B. Procedural History

Appellants sued HP for damages caused by the fire, alleging the printer was defectively designed, manufactured, and marketed and that it caused the fire. Appellants also alleged HP was negligent in designing, manufacturing, and marketing the printer.

After discovery was completed, HP filed a combined traditional and no-evidence motion for summary judgment on appellants’ products liability and negligence claims. In its motion, HP contended appellants had no evidence: (1) of a defect in the printer or that it caused the fire; (2) that the fire would normally not have occurred in the absence of negligence; and (3) that the instrumentality causing the injury was under HP’s control. In response to HP’s no-evidence summary judgment motion, appellants presented the affidavit of Otte and the depositions of their two experts on fire origin and cause: Captain Martinez of the Dallas Fire Department and Edward Roberts, a private fire origin and cause investigator.

HP’s motion also asserted that its summary judgment evidence proved as a matter of law the printer could not have caused the fire. In support of its traditional motion for summary judgment, HP presented the affidavit and report of its electrical engineering expert, Donald Galler; portions of the depositions of appellants’ two fire origin and cause experts (Martinez and Roberts); and portions of the deposition of appellants’ electrical engineering expert, Glenn Hardin.

The trial court granted HP’s motion for summary judgment without specifying the grounds therefor. Appellants timely perfected their appeal.

III. PRELIMINARY MATTERS

To simplify the discussion of the principal issues presented, we first dispose of appellants’ negligence and design and marketing defect claims.

In addition to their products liability claims, appellants alleged that HP was negligent in its design, manufacturing, and marketing of the printer, and pled reliance on the doctrine of res ipsa loquitur. Although HP’s motion directly addressed whether res ipsa loquitur was applicable to prove negligence, it did not otherwise address appellants’ negligence claims. However, the trial court’s summary judgment disposed of appellants’ negligence claims as well as their products liability claims.

On appeal, appellants present two issues; only the second issue can be said to relate to appellants’ negligence claims. In that issue, appellants assert the trial court erred in “granting HP’s motion for summary judgment on appellants’ res ipsa lo-quitur claim because appellants’ summary judgment evidence raised a genuine issue of material fact with regard to whether a fire of this nature would ordinarily occur absent negligence and whether the negligence occurred while the printer was under HP’s management and control.” HP’s brief responds in kind, asserting the trial *857 court properly entered summary judgment on “appellants’ res ipsa loquitur claim.”

Referring to res ipsa loquitur as a distinct cause of action is misleading. Res ipsa loquitur is not a separate cause of action independent of a negligence cause of action; it is “simply a rule of evidence by which negligence may be inferred by the jury.” Haddock v. Arnspiger, 793 S.W.2d 948, 950 (Tex.1990). Thus, we address appellants’ second issue in connection with their negligence causes of action.

Normally, strict products liability and negligence are separate causes of action with different elements. See Ford Motor Co. v. Miles, 141 S.W.3d 809, 315 (Tex.App.-Dallas 2004, pet. denied) (considering claims for strict liability and negligence); Otis Spunkmeyer, Inc. v. Blakely, 30 S.W.3d 678, 690 (Tex.App.-Dallas 2000, no pet.) (considering claims for manufacturing defect and breach of implied warranty of merchantability). However, here appellants alleged no negligence other than conduct relating to whether the printer was unreasonably dangerous when sold. As a result, appellants’ negligence theories are encompassed and subsumed in their defective product theories, and appellants’ burden at trial would be to prove injury resulting from a product defect. See Miles, 141 S.W.3d at 315; see also Simms v. Sw. Tex. Methodist Hosp., 535 S.W.2d 192, 197 (Tex.Civ.App.-San Antonio 1976, writ ref'd n.r.e.). Therefore, any error in disposing of appellants’ negligence claims cannot have caused the rendition of an improper judgment or prevented appellants from properly presenting their case to this Court. See Tex.R.App. P. 44.1(a). Appellants’ rights to recover against HP stand or fall on the outcome of their products liability claims. We thus affirm the trial court’s judgment as to appellants’ negligence causes of action.

Appellants also alleged all three products liability theories: defective design, defective manufacturing, and defective marketing. 2 Although HP’s motion for summary judgment focused on the elements of appellants’ defective manufacturing claim, the trial court’s summary judgment disposed of all products liability claims, not just the defective manufacturing claim. However, on appeal appellants assert no issue complaining of the trial court’s disposition of their defective design and defective marketing claims. Therefore, we affirm the trial court’s summary judgment as to them. See Pat Baker Co. v. Wilson, 971 S.W.2d 447, 450 (Tex.1998) (“It is axiomatic that an appellate court cannot reverse a trial court’s judgment absent properly assigned error.”); Eichelberger v. Mulvehill, 198 S.W.3d 487, 491 n. 2 (Tex.App.-Dallas 2006, pet. denied).

Having disposed of appellants’ negligence claims and their defective design and defective marketing products liability claims, we are left with appellants’ core cause of action: a claim for damages resulting from a manufacturing defect. It is in the context of this claim that we further consider whether the trial court properly granted summary judgment.

TV. Appellants’ Manufacturing Defect Claim

A. Elements of Appellants’ Claims

Products liability imposes strict liability on the manufacturer of an unreasonably dangerous product that is a producing cause of a plaintiffs injuries. See Firestone Steel Prods. Co. v. Barajas, 927 S.W.2d 608, 613 (Tex.1996); McKisson v. Sales Affiliates, Inc., 416 S.W.2d 787, *858 788-89 (Tex.1967). The plaintiff must prove the product was defective when it left the hands of the manufacturer and that the defect was a producing cause of the plaintiffs injuries. Ridgway, 135 S.W.3d at 600. A manufacturing defect exists “when a product deviates, in its construction or quality, from the specifications or planned output in a manner that renders it unreasonably dangerous.” Id.

B. Proof of Manufacturing Defect

“Both direct and circumstantial evidence may be used to establish any material fact.” Ridgway, 135 S.W.3d at 601. However, a product consumer or user is not in a position to know the manufacturing process or how a defect might have occurred; he is ordinarily ignorant of both the product’s intricacies and its maker’s activities. See Sipes v. Gen. Motors Corp., 946 S.W.2d 143, 155 (Tex.App.-Texarkana 1997, writ denied). Thus, a plaintiff is not required to show by direct proof how the product became defective or to identify a specific engineering or structural defect. See id.; V. Mueller & Co. v. Corley, 570 S.W.2d 140, 143 (Tex.Civ.App.-Houston [1st Dist.] 1978, writ ref'd n.r.e.). If the plaintiff “has no evidence of a specific defect in the design or manufacture of the product, he may offer evidence of its malfunction as circumstantial proof of the product’s defect.” Gen. Motors Corp. v. Hopkins, 548 S.W.2d 344, 349-50 (Tex.1977), overruled in part on other grounds by Turner v. Gen. Motors Corp., 584 S.W.2d 844, 851 (Tex.1979); Sipes, 946 S.W.2d at 155. 3 A malfunction may be shown by testimony of the product’s user about the circumstances surrounding the event in question. Sipes, 946 S.W.2d at 155.

However, “[t]he inference of defect may not be drawn ... from the mere fact of a product-related accident.” Ridgway, 135 S.W.3d at 602 (quoting Restatement (Third) of Torts: Products Liability § 3 reporter’s note to cmt. D (1998)). Thus, proof of a product failure, standing alone, is not sufficient to raise a fact question as to whether the product was defective or that it was defective when it left the hands of the manufacturer. See Cooper Tire & Rubber Co. v. Mendez, 204 S.W.3d 797, 807 (Tex.2006).

The tension between these two concepts — that circumstantial evidence can be used to prove a product defect and that proof of a product failure standing alone is insufficient to raise a fact issue as to a product defect — is aptly summarized in Justice Hecht’s concurring opinion in Ridgway: “Few would question the use of circumstantial evidence to prove products liability in appropriate cases. The hard issue is not whether it can be done, but when and how.” Ridgway, 135 S.W.3d at 603 (Hecht, J., concurring). To illuminate this hard issue — when and how circumstantial evidence is sufficient to prove a products liability claim — we examine in more detail the supreme court’s opinions in Ridgway and Mendez, as well as the opinion of the Amarillo court of appeals in Turbines, Inc. v. Dardis, 1 S.W.3d 726, 732 (Tex.App.-Amarillo 1999, pet. denied).

1. Ridgway

*859 Ridgway involved a products liability claim against a pick-up truck manufacturer for damages resulting when the vehicle caught fire while being driven. The Ridg-ways, plaintiffs in the case, were the third owners of the pick-up. The pick-up’s first owner drove it for approximately 7,000 miles and installed a spotlight in the front left door frame. The second owner drove the pick-up approximately 47,000 more miles and had it repaired four times in an effort to fix a “clunking noise” that occurred during hard turns. Three of these repair's also involved the fuel system in an attempt to increase the pick-up’s mileage. The Ridgways drove the pick-up for one month and made no repairs or modifications before it caught fire. Ridgway, 135 S.W.3d at 599.

In response to the manufacturer’s motion for summary judgment, the Ridgways presented evidence that the fire occurred and an affidavit from an expert stating he suspected the fire was caused by the electrical system. However, the expert declined to eliminate all portions of the fuel system as a possible cause of the accident, conceding “the actual cause of the fire has not been determine [sic] yet.” Id. at 600.

The supreme court held that the Ridg-ways’ circumstantial evidence was no more than a scintilla of evidence as to the existence of a defect in the pick-up at the time it left the manufacturer. Id. at 602. In doing so, the court declined to decide whether Texas law was reflected in section 3 of the Restatement (Third) of Torts, 4 stating that if such an inference of product defect can be made, “it would generally apply only to new or almost new products. Such products typically have not been modified or repaired, therefore making a product defect the likely cause of an accident.” Id. at 601.

Justice Hecht’s concurring opinion, however, further explored section 3 to explain his contention that, “while Texas law would allow proof of products liability by circumstantial evidence in certain cases, [section 3] does not accurately restate Texas law.” Id. at 602 (Hecht, J., concurring). According to Justice Hecht, section 3 is modeled on the Restatement (Second) of Torts provision dealing with res ipsa loqui-tur (section 328D). 5 Id. at 603. However, “[sjections 3(a) and (b) are less strict than the parallel provisions in sections 328D(l)(a) and (b),” and the “differences in the two provisions are such that section 3 is not an analogue of section 328D but rather a kind of res ipsa B lite!” Id. at 604. Justice Hecht noted that Texas law on res ipsa loquitur is “at least as strict as section 328D” and that there was “no reason to allow an inference of products liability any more freely than an inference of negligence.” Id.

*860 Using the facts in Ridgway, Justice Hecht illustrated the difference between the section 3 of the third Restatement and Texas law:

Section 3(a) “requires only that an injury-causing incident be the kind of thing that ordinarily results from a product defect, not that the incident is the kind of thing that ordinarily does not result unless there is a defect. A pickup suddenly bursting into flames for no discernible reason is the kind of thing that ordinarily occurs as a result of a product defect in the sense that product defects do cause such things. Thus Ridgway has satisfied section 3(a), even though it is also true that fires in vehicles ordinarily occur for many reasons other than product defect.”

Id. at 603 (emphasis added). However, under section 328D (and Texas law), the Ridgways had not shown that fires in pickups do not ordinarily occur absent a product defect; as Justice Hecht noted, “they ordinarily occur for all sorts of reasons.” Id. at 604. Nor had the Ridgways eliminated by evidence the existence of other possible causes of the fire; given the circumstances, they had not shown “another cause was impossible or even improbable.” Id.

Justice Hecht concluded with the following statement as to when circumstantial evidence is legally sufficient under Texas law to prove a products liability claim:

An inference of products liability is really two inferences: that the product was defective, and that the defect existed at the time of sale. Applying the principle underlying res ipsa loquitur, neither inference [regarding products liability] can be drawn without evidence that the injury would not ordinarily have occurred absent a product defect and that the defect probably existed when the product was sold.

Id. (emphasis added).

2. Mendez

In Mendez, the plaintiffs sued a tire manufacturer for damages resulting from a single-vehicle traffic accident. Mendez, 204 S.W.3d at 799. The plaintiffs alleged the accident occurred when a tire lost its tread as a result of a manufacturing defect. After concluding the trial court erred in admitting the testimony of the plaintiffs’ experts, the supreme court addressed whether circumstantial evidence in the record could sustain the jury’s finding of a manufacturing defect. After noting that a defect cannot be inferred from the mere fact of a product-related incident, citing Ridgway and the reporter’s note to comment D of section 3, the supreme court explained:

The mere fact that the tire failed would amount to evidence of a manufacturing defect “so slight as to make any inference a guess [and] is in legal effect no evidence.” [Ridgway, 135 S.W.3d] at 601. As we discussed in [General Motors Corp. v.] Hopkins, circumstantial evidence of a product defect may be offered, but where, in another case, “[t]he record contained no proof of the [product’s] defect except the malfunction itself,” and the product had been in use for years and subjected to many adjustments and changes, the cause of the product failure and proof of original defect “could not be answered except by speculation.” [Hopkins,] 648 S.W.2d at 349-50.

Mendez, 204 S.W.3d at 807 (first brackets original). 6

The court also adjudged legally insufficient the plaintiffs’ expert testimony at *861 tempting to eliminate other causes of the tire failure, stating “[t]he universe of possible causes for the tire failure is simply too large and too uncertain to allow an expert to prove a manufacturing defect merely by the process of elimination.” Id. at 807-08. The court noted that the record contained (apparently undisputed) evidence that: the tire at issue had 30,000 miles on it; it had a hole from a nail that had penetrated completely through the tire; and two of the plaintiffs’ experts admitted a nail hole or other puncture can cause tread separation. The court concluded: “In these circumstances we hold that plaintiffs’ attempts to eliminate other possible causes for the tire failure were legally insufficient to establish a manufacturing defect.” Id. at 808.

3. Turbines, Inc.

Turbines, Inc. involved a products liability claim against the seller of a used aircraft engine. The plaintiff, Dardis, and his company retrofitted crop dusting aircraft with turbine engines to increase their performance. Turbines, Inc., 1 S.W.3d at 730. Dardis obtained a used Pratt & Whitney turbine engine from Turbines, Inc., performed the retrofit, and was flying the plane back to his customer when the plane stalled and crashed shortly after takeoff, injuring Dardis.

According to Dardis, the engine lost power as he was performing a standard maneuver. Id. Dardis’s expert was able to rule out other systems, such as the fuel system, as the source of any failure, but found that the engine’s compressor bleed valve was missing after the crash. Id. at 732. Because he could not examine the valve and had excluded other malfunctions in other systems, the expert concluded the engine lost power as a result of a failure in the bleed valve. Id. The crash occurred in 1994; the evidence indicated the valve was on the engine from at least 1968, and possibly as early as 1964 when the engine was built. There was also evidence the engine had been used in a mechanics school for several years and had only 900 hours of flight time, a very low number of hours.

Turbines, Inc. argued the crash was not caused by mechanical failure but by pilot error. Id. at 730. Several experienced pilots who witnessed the crash testified Dardis used a very short ground roll before takeoff, climbed at a very steep angle, and made a 180-degree turn with a high bank angle. These maneuvers significantly increased the airspeed necessary to avoid a stall. The pilots testified the plane stalled and crashed into the hangar. They agreed there was no change in the engine sound during the flight. Id. at 731. The trial court entered a judgment in favor of Dardis based on the jury’s findings of liability.

The court of appeals noted that whether the engine failed was a heavily disputed issue, unlike the situation in Sipes, where it was undisputed the airbag did not activate. Id. at 736. 7 Dardis’s argument, and his expert’s conclusion, began with the assumption that the engine failed. The only evidence that the bleed valve failed was “the inference made [by Dardis’s expert] that the bleed valve must have failed because he excluded the possibility of malfunction of other engine systems.” Id. This inference was then used as circumstantial evidence that there was a defect of some unknown type in the bleed valve. 8 *862 The court rejected Dardis’s product defect theory not because of the circumstantial nature of the evidence, but because the theory required the use of one inference as circumstantial evidence to support another inference to establish the ultimate fact. Id. “It is well established that facts may not be found based on chains of inferences.” Id. The court concluded there was no evidence of a defect in the design, manufacture, or marketing of the engine. Id. at 737.

4. Summary

From the supreme court’s opinions in Ridgway, Mendez, and the other supreme court opinions cited herein, and the logic underlying Justice Hecht’s concurring opinion in Ridgway, the Amarillo court’s opinion in Turbines, Inc., and the other intermediate appellate court opinions cited herein, we make the following observations with respect to when evidence is sufficient to avoid summary judgment on a manufacturing defect products liability claim.

First, the evidence is sufficient to avoid summary judgment when, viewing all the evidence in the light most favorable to the non-movant and indulging every reasonable inference and resolving any doubts against the motion, more than a scintilla of probative evidence exists as to whether: (1) the product was defective when it left the hands of the manufacturer; and (2) the defect was a producing cause of the plaintiffs injuries. Such evidence must do more than create a mere surmise or suspicion with respect to these two elements. See Sudan, 199 S.W.3d at 292; Ridgway, 135 S.W.3d at 600; King Ranch, Inc., 118 S.W.3d at 751; Firestone Steel Prods. Co., 927 S.W.2d at 613. Thus, if the evidence is such as to render any inference of these elements no more than a guess, it is insufficient. Ridgway, 135 S.W.3d at 601; see also Wal-Mart Stores, Inc. v. Gonzalez, 968 S.W.2d 934, 936 (Tex.1998) (circumstantial evidence is legally insufficient if it admits of two equally plausible but opposite inferences because selecting one inference over the other is speculation). This is merely a restatement of the summary judgment standard of review in the specific context of a manufacturing defect claim.

Second, “[b]oth direct and circumstantial evidence may be used to establish any material fact.” Ridgway, 135 S.W.3d at 601. Thus, the “more than a scintilla” of evidence necessary to avoid summary judgment on a manufacturing defects claim can be supplied through direct evidence, circumstantial evidence, or a combination of both.

Third, however, for circumstantial evidence to support an inference that the product was defective, it must do more than raise the possibility the injury could have resulted from a defect. This is because the product’s failure, standing alone, provides no more than a basis for guessing as to whether the product was defective or whether it failed as a result of other causes. See id. at 601-02 (vehicle fire); see also Mendez, 204 S.W.3d at 807 (tire failure). 9 For circumstantial evidence to support an inference that the product *863 was defective, it must provide a reasonable basis for concluding the injury would not ordinarily have occurred absent a defect. See Ridgway, 135 S.W.3d at 604 (Hecht, J., concurring).

Fourth (and similar to the third observation), for circumstantial evidence to support an inference that the defect existed when it left the manufacturer, it must do more than raise the possibility the defective condition could have existed at that time. Again, this is because evidence that the product failed, standing alone, provides no more than a basis for guessing as to whether the product was defective when it left the manufacturer’s control. For the circumstantial evidence to support such an inference, it must provide a reasonable basis for concluding the defective condition did not arise subsequent to the manufacturer’s exercise of control over the product. See id.; Hopkins, 548 S.W.2d at 350.

Sometimes this basis can be provided by evidence of the age of the product and its history of usage up to the time of failure. “The age and use of [a] product during the time intervening between [its] purchase and malfunction will tend to support or defeat the circumstantial weight of the malfunction as proof of original defect.” Hopkins, 548 S.W.2d at 350. New or nearly new products “typically have not been modified or repaired, therefore making a product defect the likely cause of an accident.” Ridgway, 135 S.W.3d at 601. Thus, an inference of original product defect may be warranted from the malfunction of a relatively new or sealed product. See Darryl, 440 S.W.2d at 631 (three-month-old truck with 600 to 700 miles and no repairs); Sipes, 946 S.W.2d at 146, 155 (sealed airbag system in new car). However, such an inference may not be warranted if the product was worn, misused, damaged, repaired, or altered after it left the manufacturer’s control. See Mendez, 204 S.W.3d at 808 (tire with 30,-000 miles and nail puncture); Ridgway, 135 S.W.3d at 599 (two-year-old truck with 54,000 miles and three repairs to fuel system); Turbines, Inc., 1 S.W.3d at 732 (allegedly defective valve in 30-year-old aircraft engine).

Fifth, for circumstantial evidence to support inferences that a product was defective and the defect existed at the time it left the manufacturer, the evidence need not disprove all other possible causes for the injury. See Darryl, 440 S.W.2d at 632; Parsons v. Ford Motor Co., 85 S.W.3d 323, 331 (Tex.App.-Austin 2002, pet. denied). For example, “[t]he plaintiff is not required to exclude an appreciable chance that the event might have occurred in some other way. Expressed otherwise, a conclusion of causal connection may be inferred by a balance of probabilities.” Sharp v. Chrysler Corp., 432 S.W.2d 131, 135 (Tex.Civ.App.-Houston [14th Dist.] 1968, writ ref'd n.r.e.) (product case). However, the likelihood óf other possible causes must be so reduced that the fact-finder could reasonably find by a preponderance of the evidence that the cause of the product failure lies at the manufacturer’s door. See Ridgway, 135 S.W.3d at 604 (Hecht, J., concurring) (citing Mobil Chem. Co. v. Bell, 517 S.W.2d 245, 251 (Tex.1974) (negligence case)); see also Merrell Dow Pharms., 953 S.W.2d at 720 (in toxic tort products case, evidence must exclude other plausible causes of the injury or condition “with reasonable certainty”).

Sixth, a number of inferences may be drawn from a single fact situation. See McClure v. Allied Stores of Tex., Inc., 608 S.W.2d 901, 904 (Tex.1980) (citing Farley v. M M Cattle Co., 529 S.W.2d 751, 756 (Tex.1975) (proposition that a number of inferences may be drawn from single fact situation is “well established”)). Thus, circumstantial evidence can give rise to both *864 the inference of a product defect and the inference that the defect existed at the time of sale. However, “an inference stacked only on other inferences is not legally sufficient evidence.” Marathon Corp. v. Pitzner, 106 S.W.3d 724, 728 (Tex.2008) (citing Lozano v. Lozano, 52 S.W.3d 141, 148 (Tex.2001)). 10 This principle is illustrated in Turbines, Inc., in which the plaintiff used one inference (that the engine’s bleed valve must have failed because the expert excluded the possibility of malfunction in the other engine systems) as the sole circumstantial evidence supporting another inference (that there was a defect of an unknown type in the bleed valve). Turbines, Inc., 1 S.W.3d at 736.

With these principles in mind, we turn to the grounds for summary judgment presented in this case.

Y. Summary Judgment Grounds

A. No Evidence Motion

1. Summary Judgment Evidence

In response to the motion for summary judgment, appellants presented Otte’s affidavit and the depositions of two fire origin and cause experts: Captain Martinez of the Dallas Fire Department and Edward Roberts, a private fire origin and cause investigator.

According to Otte's affidavit, the printer was new, arrived in a sealed box, and appeared undamaged. He set up and used the printer according to HP’s instructions for approximately two months without problem. The printer was not damaged, repaired, or modified before the fire. The printer was located on the left end of the credenza in Otte’s office and plugged into a power strip. The printer’s power supply was on the floor beneath the printer. (According to HP’s expert, the printer’s power supply was never found after the fire.) The printer was the only electrical device on the credenza that was plugged in, and the printer’s power supply was the only electrical device at floor level near the printer and credenza. Otte stated the printer was severely damaged in the fire.

Captain Martinez is a trained arson and fire investigator; at the time of the fire he had four years’ experience as a fire investigator. According to his deposition, Martinez investigated the fire scene the day after the fire during salvage operations. He interviewed firefighters and other people, including Otte, about the fire, the arrangement of the office, and the location of the furniture and equipment. Martinez said Otte described everything that was on the desk and surrounding tables and did not indicate anything was missing. From interviewing Otte, Martinez understood how the office equipment was placed in the office and reconstructed the scene accord- *865 mg to where Otte said the equipment was positioned.

Martinez determined the area of origin by going from the area of least burn damage to the area of most burn damage. He concluded the origin was toward the window by a desk or credenza. The amount of fire damage to the credenza and the items on the credenza indicated the origin of the fire. The printer was more damaged than the other pieces of equipment he examined. The greater amount of fire damage to the printer indicated it burned longer than other items, leading him to conclude that was where the fire started. He determined that some type of electrical malfunction in the printer caused the fire, although he admitted he was not an electrician and could not say whether the printer was defective. Martinez also determined the fire damage to the other computer equipment in the office was due to exposure from another source.

Roberts was a private fire origin and cause investigator with eight years’ experience at the time of his deposition. He inspected the site shortly after the fire and retrieved several bags of debris from the dumpster. Roberts interviewed Otte and Martinez about the fire and the layout of the office before the fire. There was a large desk in front of the credenza and a small table between the credenza and desk on the north wall. 11 The credenza was located on the east wall near the northeast corner of the office. The printer was located on the north end of the credenza. Photographs 12 of the office indicated the faceplate of the electrical outlet (duplex receptacle) on the north wall was undamaged.

Roberts did not see the credenza after the fire, but viewed photographs of the fire damage to the credenza and discussed it with Martinez and Otte. The northern end and back of the credenza were significantly damaged. Photographs showed the northern end of the credenza had fallen over. In contrast, there was fire damage to the right-hand drawers of the desk (in front of the left end of the credenza), but a lack of fire damage underneath the desk.

Based on the burn patterns in the office and the evidence that the credenza and table were more damaged than the desk, Roberts determined the fire originated in the northeast corner. He also determined the fire did not originate on the floor because the tops of the desk and credenza were more damaged than the bottoms.

Two computer monitors were in the office, one on the large desk and the other on the table between the desk and credenza. Roberts examined a computer monitor and keyboard that had been located on the table between the desk and credenza. The underside of both were relatively undamaged and the damage to them had directional patterns indicating the fire spread to them from another source (rather than originating in them). 13

After determining the point of origin, Roberts looked at the potential ignition *866 sources in the area. He said, “In this case, the only thing that was there was this printer that had been more heavily damaged than anything else in the room....” The printer “had much more damage to it and it’s the only heat producing item in the area of origin. Nothing else was plugged in over there.” He explained that the printer was “by far more damaged than anything there, than the CPUs, than the monitor, than the keyboard, than any of the electrical items associated with the structure, duplex receptacles things of that nature.”

Roberts explained his conclusion that the printer was the only ignition source at the point of origin of the fire:

Q. And you’re not even saying it’s possible then that the printer was the cause of this accident. You’re just saying that that’s where you determined that the origin was and the printer happened to be there; is that right?

A. The printer is the only possible ignition source there, yes.

Q. But you don’t even know that the printer is a possible ignition source, do you?

A. Yes, I do.

Q. Why?

A. It was plugged in at the time, located right at the point of origin and no other ignition source is there.

Q. Well, you don’t know that just being plugged in can make it a source of ignition, do you?

A. Well the fact that its [sic] burned and it started there indicates to me that it can.

Roberts was questioned about a letter report from David Sneed concerning the origin of the fire. (Sneed’s letter is not in the summary judgment record; however, Roberts testified as to why he disagreed with the letter’s conclusion that the monitor could have been a source of the fire or that the fire originated at or near the floor level.) The computer equipment on the floor at that place had insignificant damage and the underside of the monitor and keyboard on top of the table were also undamaged. Roberts testified that the lack of damage to the underside of the monitor and keyboard and to the CPUs and the more extensive damage to the printer and credenza cannot be explained by Sneed’s theory that the fire originated with the monitor or at or near floor level.

HP asserts that appellants’ engineering expert, Glenn Hardin, testified he was unable to find any electrical defect in the printer. However, this statement is not in the summary judgment evidence. Hardin did say he was unable to form an opinion about whether or not the printer caused the fire. He also testified the printer “had the capability of starting the fire. And as to whether or not it actually caused the fire, I’m not prepared to make that statement.” He concluded the printer had power going to it and “Components can fail inside electronic products. Can and do fail.”

The record also contains the affidavit and report of HP’s electrical engineer, Gal-ler. This evidence is described in more detail in the section on the traditional motion of summary judgment, but briefly, Galler concluded from examining the remains of the printer and other evidence that appellants’ experts had not scientifically proven the printer could have caused the fire, the printer did not cause the fire, damage to the computer monitor was consistent with the fire origin reported by Sneed, and an electrical failure in the monitor or its power cord could have been a cause of the fire.

2. Application

Viewing all the summary judgment evidence in the light most favorable to appellants — including the expert evidence as to *867 cause and origin of the fire and the circumstantial evidence upon which that testimony is based — reasonable fact-finders could differ in their conclusions as to whether the fire started in the printer. Office printers — especially new or nearly new printers that have not been damaged or abused — do not ordinarily catch fire absent a product defect. The record further shows the printer was new and had not been altered or repaired by third parties. From this and the other evidence in the record, reasonable fact-finders could differ in their conclusions as to whether the printer was defective and that the defect probably existed at the time it left the manufacturer. See Ridgway, 135 S.W.3d at 604 (Hecht, J., concurring); see also Redman Homes, Inc. v. Ivy, 920 S.W.2d 664, 668 (Tex.1996) (concluding in breach of warranty case that circumstantial evidence from fire origin and cause expert was some evidence of defect in wiring of mobile home). Thus, reasonable minds could differ in their conclusions as to whether the printer was defective when it left the hands of the manufacturer and that the defect was a producing cause of appellants’ injuries.

HP argues appellants did not exclude all other possible ignition sources. However, both Martinez and Roberts testified that, in their opinion, the fire started on the credenza where the printer was located and gave the reasons supporting their opinions. There is also evidence the printer was the only device located on the credenza that was plugged in. We conclude this is sufficient evidence from which reasonable fact-finders might reach differing conclusions as to whether the fire started in the printer.

3. Summary

Reviewing the summary judgment evidence in the light most favorable to appellants, we conclude reasonable minds could differ in them conclusions as to whether the printer was defective at the time it left HP’s control and whether it was a producing cause of appellants’s injury. Accordingly, the trial court erred in granting summary judgment based on the no-evidence motion.

B. Traditional Motion

HP’s traditional motion for summary judgment was based on the affidavit and report of its electrical engineering expert, Galler, which HP contends conclusively establish that the printer did not cause the fire.

Galler is an electrical engineer with more than fifteen years’ experience in failure analysis including electrical failure analysis. He does not purport to be an expert in the investigation of the origin and cause of fires. Galler reviewed the reports and depositions of Hardin, Roberts, and Martinez. He also reviewed x-rays of the damaged printer, schematics of the printer, the Underwriter’s Laboratory report on the HP model printer, and examined and tested an exemplar model printer and AC adapter. He later examined the remains of the printer and of the computer monitor and keyboard. Galler expressed the following opinions: (1) plaintiffs experts have not scientifically proven that it is possible for the printer to cause a fire under any circumstances; (2) the printer did not cause the fire; (3) the damage to the monitor power cord and enclosure is consistent with the fire origin reported by Sneed; and (4) an electrical failure of the monitor or the monitor power cord could have been a cause of the fire.

Galler determined the printer cartridges were in the parked position, indicating the machine was in idle mode at the time of the fire. He examined the printed circuit board in the printer and found it to be heavily charred on the component side with only a few components remaining. *868 The damage was uniform and the bottom of the circuit board was less damaged than the top. He would expect the portions of the circuit board to be completely destroyed if the printer had been the cause of the fire because the circuit board is where the eighteen-volt direct current power is connected and distributed to the motors in the printer. Galler noted the plastic enclosure of the printer was completely consumed in the fire, but the paper tray had a stack of paper about three-eights of an inch thick. The paper was charred around the edges, but otherwise unburned. Stacked paper does not burn as well as loose paper, but Galler was surprised so much paper survived the fire if the fire originated in the printer. Galler also observed the printer’s AC power adapter was not found. His report indicates the monitor was damaged on the side and top, but the base was still intact. He examined a power cord that matched the remains of a power cord which had been connected to the monitor. The cord insulation was intact except near the end where several small copper-colored beads were found indicating some arcing had taken place. Galler stated the beading could have been caused by a fire attack damaging the insulation while the monitor was energized or by a failure in the cord due to mechanical stress, resulting in heating and arcing. This arcing could have been an initiating event in the fire origin.

In contrast, there is evidence the fire originated at the place the printer was located and the printer was the only device plugged into an electrical source at that location. The printer was the most heavily damaged piece of equipment in the office. Roberts explained that Sneed’s report is not consistent with the location of the fire origin and Sneed did not account for the greater damage to the printer and the lack of damage to the CPUs and the underside of the monitor and keyboard. There is no evidence of mechanical stress to the power cord other than Galler’s speculation. Gal-ler was not an expert in determining fire origin and cause, and he did not explain how the fire could have originated in the printer without a product defect. We also note that Sneed’s experience, methodology, and actual opinions on the origin of the fire are not in the summary judgment record.

Galler’s opinion that the printer did not cause the fire is some evidence that, if believed, could negate appellants’ product liability claim. However, HP’s burden on a traditional motion for summary judgment is to negate an essential element of appellants’ cause of action as a matter of law. Friendswood Dev. Co., 926 S.W.2d at 282. As discussed above, there is evidence in the record from which reasonable fact-finders could differ in their conclusions as to whether the printer was defective, the defect existed when it left the manufacturer, and the defective printer caused the fire and appellants’ injury. In view of the other evidence in the record, Galler’s opinion does not negate as a matter of law an essential element of appellants’ manufacturing defect claim.

After reviewing the summary judgment evidence in the light most favorable to appellants, we conclude HP did not meet its summary judgment burden of negating as a matter of law an essential element of appellants’ cause of action.

VI. Conclusion

Based on the summary judgment evidence, we conclude there is a genuine issue of material fact as to whether the printer was defective when it left the manufacturer and caused the fire. We also conclude that HP did not negate an essential element of appellants’ claims as a matter of law. Accordingly, we sustain appellants’ issues regarding the manufacturing defect claims. We affirm the trial court’s summary judgment as to appellants’ negli- *869 genee and design and marketing defect claims. We reverse the trial court’s summary judgment on the manufacturing defect claim and remand the case for further proceedings.

. In turn, Merrell Dow Pharmaceuticals, Inc. cites Robert W. Calvert, "No Evidence" and “Insufficient Evidence" Points of Error, 38 Tex. L.Rev. 361, 362-63 (1960).

. A product may be unreasonably dangerous because of a defect in marketing, design, or manufacturing. Am. Tobacco Co. v. Grinnell, 951 S.W.2d 420, 426 (Tex.1997).

. Indeed, in a products liability case circumstantial evidence will often be the only available evidence of a defect:

It would be equally difficult, if not impossible, for the plaintiff to rebut by direct evidence all of the conceivable possibilities which would account for the defective condition other than the existence of the defect at the lime of the sale. Such direct evidence should not be required, particularly when dealing with a latent defect.

See Darryl v. Ford Motor Co., 440 S.W.2d 630, 632 (Tex.1969) (“To exclude circumstantial evidence that the product was defective at the time of the sale would frustrate the beneficial purposes of the [strict liability] doctrine.”).

. Section 3 states:

It may be inferred that the harm sustained by the plaintiff was caused by a product defect existing at the time of sale or distribution, without proof of a specific defect, when the incident that harmed the plaintiff:

(a) was of the kind that ordinarily occurs as a result of a product defect; and

(b) was not, in the particular case, solely the result of causes other than the product defect existing at the time of sale or distribution.

Restatement (Third) of Torts: Products Liability 3 (1998).

. Section 328D states:

(1) It may be inferred that harm suffered by the plaintiff is caused by negligence of the defendant when

(a) the event is of a kind which ordinarily does not occur in the absence of negligence;

(b) other responsible causes, including the conduct of the plaintiff and third persons, are sufficiently eliminated by evidence; and

(c) the indicated negligence is within the scope of the defendant’s duty to the plaintiff.

Restatement (Second) of Torts § 328D (1965).

. The court also held the mere fact that the tire failed was insufficient to establish a manufacturing defect because the failure could have been caused by a design defect — a theory not tried and about which the jury was not asked. Id. at 799, 807, 808.

. The Turbines, Inc. opinion mistakenly refers to the Hopkins case as the one where it was undisputed the vehicle airbag did not activate. Hopkins involved the defective design of a carburetor, 548 S.W.2d at 346-47, while Sipes involved an alleged defect in the airbag system of an automobile. 946 S.W.2d at 146, 156.

. Yet, the same expert stated the failure was due to age and "there was no defect in the *862 design, manufacture, or marketing of the component." Turbines, Inc., 1 S.W.3d at 736.

. See also Mack Trucks, Inc. v. Tamez, 206 S.W.3d 572, 583-84 (Tex.2006) (circumstantial evidence that fire quickly reached driver following truck rollover, although consistent with plaintiffs' theory that fire originated from allegedly defective fuel system and location of ignition sources, "does not make it more likely than not that the battery or some other allegedly improperly located ignition source ignited diesel from the tractor, as opposed to other possible sources of ignition such as the cargo of crude oil”); Gen. Motors Corp. v. Iracheta, 161 S.W.3d 462, 470-72 (Tex.2005) (mere possibility that fire in truck occurred in manner plaintiff suggested was not enough to support jury's findings).

. This “purported rule” has been criticized: "Actually, that is not a separate rule of decision: it is also but another way of saying that the vital fact may not reasonably be inferred from the meager facts proved in the particular case.” Robert W. Calvert, “No Evidence" and “Insufficient Evidence” Points of Error, 38 Tex. L.Rev. 361, 365 (1960); see also Phoenix Ref. Co. v. Powell, 251 S.W.2d 892, 902 (Tex.Civ.App.-San Antonio 1952, writ ref. n.r.e.) ("This form of statement is often employed by the courts, and has been used by this writer, as and for a shorthand rendition of the rule which requires that verdicts be based upon more than surmise and guesswork. The rule can not be literally applied in all factual circumstances.”); William V. Dor-SANEO, III ET AL„ TEXAS LlTICATION GUIDE § 146.03[6][e][ii][B] (online ed. 2007) ("Texas courts would be much better off to abandon the inference-piling approach to "no evidence” review because it is hard to apply, easy to manipulate, and simply not a reliable tool for assessing the reasonableness of an inference.”); 1A John H. Wigmore, Evidence § 41 (Tillers Rev. 1983) (“There is no such orthodox rule; nor can there be. If there were, hardly a single trial could be adequately prosecuted.”).

. The deposition testimony references diagrams of the office, but these diagrams were not included in the summary judgment record.

. The actual photographs mentioned in the deposition testimony are not included in the summary judgment record.

. Apparently the photographs of the monitor and keyboard did not show one side was more damaged than the other. However, based on his examination at the time, it appeared to Roberts that the direction of fire travel was from the east to the monitor. HP’s expert, Donald Galler, an electrical engineer, also noted the monitor was more damaged on one side. Roberts's deposition refers to a photograph showing the power cord to the monitor located on the desk, but does not describe the condition of this cord.