571 So.2d 140 (1990)
Ronald REILLY
v.
DYNAMIC EXPLORATION, INC., et al.
William O. BARRETT, Sr.
v.
PRODUCTION MANAGEMENT CORPORATION, INC., et al.
Nos. 90 C 0638, 90 C 0681.
Supreme Court of Louisiana.
December 3, 1990.
Warren A. Perrin, Gerald C. deLaunay, Lafayette, for William O. Barrett, Sr., plaintiff-applicant.
Paula Cobb, Arthur Cobb, Cobb & Cobb, Baton Rouge, for Ronald Reilly, plaintiff-applicant.
Charles A. O'Brien, Franklin, Moores & Walsh, Baton Rouge, for Kimray Inc., defendant-respondent.
Stephen R. Wilson, Keogh, Cox & Wilson, Baton Rouge, for Stockham Valves & Fittings, defendant-respondent.
WATSON, Justice.
After a jury trial of these suits for personal injuries resulting from an explosive fire at an oil well, the trial judge granted directed verdicts to defendants, Stockham Valves & Fittings, Inc. and Kimray, Inc. The court of appeal affirmed.[1] A writ was granted to review the judgment of the court of appeal.[2]
Since the trial court granted directed verdicts to defendants, the evidence is properly summarized in the light most favorable to plaintiffs. Breithaupt v. Sellers, 390 So.2d 870 (La.1980). See Boeing Company v. Shipman, 411 F.2d 365 (5th Cir.1969).
FACTS
On May 4, 1981, plaintiffs, William O. Barrett, Sr. and Ronald B. Reilly, were working on the No. 2 well owned by Dynamic Exploration, Inc. at Mallet's Bluff, East Baton Rouge Parish, Louisiana. Plaintiffs' theory of recovery against these defendants was that the explosion and injuries were caused by a valve manufactured by Kimray and/or a check valve manufactured by Stockham, which were defective, resulting in a backflow of gas and an explosive fire.
Reilly was employed as a gauger by Production Management, which had been managing the Dynamic facility at Mallet's Bluff. Barrett was supervising a crew from Central Industries which was hired to perform a bypass of a separator at the facility. Reilly described the location as dangerous. Among other deficiencies, a schematic diagram of the piping system was not available.
Management of the wells was in the process of being transferred from Production Management to Dynamic. Don Dronet of Dynamic, who was taking over management of the wells at Mallet's Bluff, confirmed Reilly's testimony that conditions at the field were bad and described them as about the worst he had seen in twenty years.
Reilly and Barrett were attempting to bypass the high pressure separator at the No. 2 well. The oil flows from the No. 2 well through a choke, then through a high pressure separator, on to a low pressure separator, and then through a heater treater. At the heater treater, heat is applied to separate the gas from oil and water. Gas then goes through a Kimray valve and into the ground where a two inch Stockham check valve is buried six inches below the surface. Dronet testified that he would have hooked up the Kimray valve with a blocking valve union on each side so that the valve could be isolated. Several experts said that the Stockham valve required frequent maintenance to function properly and should not have been buried beyond easy access.[3] The gas went into a common manifold or pipeline. All four heater treaters at the site were hooked into that manifold. Excess gas from the four wells went eighty feet through a two inch overflow line to an open pit for evaporation.
Before attempting the bypass, Reilly, accompanied by Barrett, closed several valves. They bled down the high pressure separator and the low pressure separator and relieved the valve upstream from the high pressure separator. Since they still had pressure on the line, they rechecked their procedure. They tested the gauge on the Kimray valve, which registered 17 to 25 psi, but it was accurate.
An arrow on the Kimray valve pointed downstream into the ground. There was extensive testimony that an arrow indicates a one way flow in the arrow's direction. After an unsuccessful attempt to break the pipe union downstream, Barrett and Reilly decided to break the pipe union upstream between the heater treater and the Kimray valve to dissipate the pressure. They did not know there was a check valve in the ground and relied on the Kimray valve to act as a check valve on the upstream pressure. The Kimray valve resembles a check valve and generally prevents a reverse flow of gas.
Reilly and Barrett cut off the gas to the burner and the pilot light on the heater treater and checked that the fires were out. Since 17 to 25 psi is a small amount of pressure, they expected to be able to bleed it off in three or four minutes. When they cracked the union, there was a hissing of gas which stopped. After the gas flow ceased, Barrett parted the union. He stuck out his hand, felt a surge through the Kimray valve and an almost instantaneous explosion which covered him with fire.
The fire extended at least twenty feet from the Kimray valve and lasted for approximately twenty minutes. The residual flame also came from the Kimray valve. Barrett was severely injured and Reilly's arms were burned. The accident apparently occurred because the gas backed upstream from the Kimray and Stockham valves.
Paul Montgomery, an expert in the field of petroleum engineering, testified that the purpose of the Stockham check valve was to make the gas flow in only one direction. If a check valve is functioning properly, it is impossible for gas going through the valve to flow back in the opposite direction. According to Montgomery, the length and direction of the fire indicated a defect in the check valve. If the valve had held, the fire would have lasted only two or three minutes; therefore, the gas must have backed up from the valve. In Montgomery's opinion, this explosive fire would not have occurred if the Stockham valve had worked.
Montgomery said several things combined to cause the accident: the presence of natural gas; the fire box in the heater treater; lack of a proper schematic diagram; a malfunction in the check valve; and inadequate supervision. Subsequent to this accident, a flame arrester was installed on the back of the heater treater at the No. 2 well.
Wiley D. Poole, an expert in mechanical engineering, investigated the accident on behalf of Liberty Mutual Insurance Company, the compensation insurer. Poole and an insurance adjuster visited the scene on May 5, 1981, the day after the accident. The broken connection had been put back together but the line was not operating. From Poole's observation, the fire was so great that its main source had to be the manifold and "the Kimray valve was feeding backwards." (Joint Testimony, Vol. 2, p. 38.)
When the Kimray valve was disassembled, the valve cover was sealed closed. In Poole's opinion, this resulted from the fire. Poole did not find anything unreasonably dangerous in the design of the Kimray valve but he thought it needed a check valve downstream producing zero pressure in order to operate properly.
On Poole's second visit to the scene, the gas line had been uncovered and he was able to photograph the Stockham valve. That valve does not regulate flow; it allows gas to pass in only one direction, the direction of the arrow. In July of 1984, Poole tested the Stockham valve with an air compressor to determine the backflow of air through the valve. The first test of 20 psi went right through the valve in the opposite direction from the arrow. After repeated testing, the valve held a little longer but it leaked every time it was tested. The increase in function as the tests progressed was not surprising because trash got blown out by the air pressure. Poole had no doubt that the source of the fire's fuel was the Stockham valve.
Robert D. Owen, Sr., an independent safety consultant, was tendered as an expert in work place safety. The trial court sustained an objection to his testimony on the ground that this field of expertise was not sufficiently defined; however, Owen was allowed to testify subject to contemporaneous objections. Owen said the procedure followed by Barrett and Reilly was reasonable for a line with some pressure. In Owen's experience, check valves are reliable for preventing backflow of fluids but not for any type of gas. In connection with his testimony, it was stipulated that, if it were necessary to shut off the flow of gas, a check valve should be used in conjunction with a manual gate valve.
Thomas A. Hill, Executive Vice President and General Manager of Kimray, testified that the Kimray valve does a good job of preventing pressure from going backwards but is not fabricated for that purpose. It is designed to control pressure but does not work unless the upstream pressure is higher than the downstream pressure. Under normal conditions, the valve does not allow reverse flow, but a reverse flow could have occurred if debris had propped the valve open. If more than one Kimray valve were connected to a manifold line, each should be isolated with check valves, standard procedure in the oil patch. The Kimray literature does not warn users it is not reliable as a check valve, but Hill said it is hazardous to rely on the Kimray valve as a check valve.
Hill admitted that: if there were no gas upstream, no restriction in flow downstream, the gauge failed to reduce, and the gauge was working, then the Kimray valve may have failed. Thus, if the jury had accepted Reilly and Barrett's account of how the accident occurred, they could have deduced from Hill's testimony that the Kimray valve failed before the fire occurred.
Calvin Rush McClinton, the Stockham representative at trial, testified that a check valve would not allow any massive backflow unless the valve were jammed open by some foreign matter. The Stockham literature does not warn against burying these valves. In 1986, Stockham ran four tests of the valve with water pressure; the valve performed as well as a new valve on the last two tests. McClinton admitted that the valve might have been cleaned by the testing procedure.[4]
At the conclusion of plaintiffs' case, defendants Stockham and Kimray moved for a directed verdict. The trial court took the motion under advisement until conclusion of the evidence, stating: "... I feel that I am in a position where I could rule on the motion for directed verdict at this time. However, because I want the record to be complete I am going to take the motion for directed verdict under advisement until the evidence is concluded." (Joint Testimony, Vol. 3, p. 189.) Defendants adopted the testimony of some earlier witnesses but did not produce any further evidence. The trial court then granted the motions for directed verdict on behalf of both defendants.
The trial judge did not give reasons for granting defendants' motions. Apparently, he concluded reasonable persons could not find that defendants' valves were unreasonably dangerous products under the criteria in Halphen v. Johns-Manville Sales Corp., 484 So.2d 110 (La.1986).[5]
LAW
A motion for directed verdict is governed by LSA-C.C.P. art. 1810, which states:
A party who moves for a directed verdict at the close of the evidence offered by an opponent may offer evidence in the event that the motion is not granted, without having reserved the right so to do and to the same extent as if the motion had not been made. A motion for a directed verdict that is not granted is not a waiver of trial by jury even though all parties to the action have moved for directed verdicts. A motion for a directed verdict shall state the specific grounds therefor. The order of the court granting a motion for a directed verdict is effective without any assent of the jury.
The standard for granting a directed verdict mirrors the standard for granting a summary judgment. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986); Hoffman v. Allied Corp., 912 F.2d 1379 (11th Cir.1990).
A directed verdict on liability is appropriate only when the trial judge is convinced that reasonable minds could not differ in their interpretation of the evidence. Scott v. Hosp. Serv. Dist. No. 1, 496 So.2d 270 (La.1986); Breithaupt, supra. See Ramos v. Liberty Mut. Ins. Co., 620 F.2d 464 (5th Cir.1980), cert. denied, 449 U.S. 1112, 101 S.Ct. 921, 66 L.Ed.2d 840 (1981). Questions of credibility should not be resolved by a directed verdict. See Henson v. Falls, 912 F.2d 977 (8th Cir. 1990) and Dalton v. Toyota Motor Sales, Inc., 703 F.2d 137 (5th Cir.1983).
The requisites for tort recovery against a manufacturer in a product liability suit are proof that: harm resulted from the product's condition; the condition made the product unreasonably dangerous in normal use; and the condition existed when the product left the manufacturer's control. Halphen v. Johns-Manville Sales Corp., 484 So.2d 110 (La.1986); Bell v. Jet Wheel Blast, 462 So.2d 166 (La.1985); Hebert v. Brazzel, 403 So.2d 1242 (La.1981); DeBattista v. Argonaut-Southwest Ins. Co., 403 So.2d 26 (La.1981); Hunt v. City Stores, 387 So.2d 585 (La.1980); Chappuis v. Sears, Roebuck & Co., 358 So.2d 926 (La. 1978); Weber v. Fidelity & Casualty Ins. Co. of New York, 259 La. 599, 250 So.2d 754 (1971).
There are four categories of unreasonably dangerous products: (1) products unreasonably dangerous per se; (2) products constructed with an unintended condition making them more dangerous than they were designed to be; (3) products with any danger inherent in normal use not known or obvious to the user of which the manufacturer has not warned; and (4) products unreasonably dangerous in design. Halphen, supra.
Normal use of a product includes any reasonably foreseeable misuse. Ingram v. Caterpillar Machinery Corp., 535 So.2d 723 (La.1988); Bloxom v. Bloxom, 512 So.2d 839 (La.1987); Hunt, supra; Chappuis, supra. A manufacturer can be liable for injury resulting from failure to warn of a danger not known or obvious to a user who engages in a reasonably foreseeable misuse. Toups v. Sears, Roebuck and Co., Inc., 507 So.2d 809 (La.1987); Winterrowd v. The Travelers Indemnity Co., 462 So.2d 639 (La.1985); Hebert, supra; Chappuis, supra; Bloxom, supra; Ingram, supra.
CONCLUSION
If injury is caused by a product which is defective because it proves unreasonably dangerous in normal use, the manufacturer is responsible if the defect existed when the product left the manufacturer's control. A jury could have concluded that defects in these valves were a cause of plaintiffs' injuries; that the fire would not have occurred if the valves were not defective in failing to regulate and/or check a massive backflow of gas; and that the defects existed when the valves left the manufacturer's control.
Although not defective, a product may be unreasonably dangerous if the manufacturer fails to adequately warn of any danger inherent in normal use which is not known or obvious to the ordinary user. Normal use includes misuse which might reasonably have been anticipated by the manufacturer.
The Stockham product guide emphasizes the importance of proper maintenance, but there is no warning against a valve being buried where it is not accessible to inspection or maintenance. There was testimony that this is an unsafe practice, which can cause clogging of the valve. Whether burial of the valve without easy access was a reasonably foreseeable misuse about which Stockham should have warned and whether any omission by Stockham contributed to the injuries were factual questions which should have been determined by the jury.
A reasonable juror could have concluded that the Kimray valve was unreasonably dangerous because it appeared to serve the function of a check valve but failed to do so. In addition, the Kimray valve may have been unreasonably dangerous because it was not isolated by blocking valves. Whether the manufacturer's warnings about proper use of this valve were adequate and whether any omission in warning contributed to plaintiffs' injuries were jury questions.
Defendants argue that plaintiffs' case is fabricated and that their account of the accident is unworthy of belief. However, plaintiffs' account was corroborated by Poole, who was at the scene the day after the accident. In any event, issues of credibility should not be decided in connection with a motion for a directed verdict.
The record is replete with disputed issues of material fact. It is impossible to say that a reasonable jury could not have resolved these issues in favor of plaintiffs. Reasonable persons could have decided for plaintiffs; reasonable persons could have decided for defendants. The trial court erred in granting the directed verdicts and depriving these plaintiffs of their right to a jury trial.
FURTHER PROCEEDINGS
This accident happened nine years ago; it would be desirable to terminate the litigation. Under Gonzales v. Xerox, 320 So.2d 163 (La.1975) this court could decide the cases in the interest of judicial economy. However, other concerns, particularly apportionment of fault among the parties, some not before the court, make a new trial the procedure of choice.
The manner in which the directed verdict was handled by the trial court undoubtedly had a chilling effect on defendants' presentation of their case; they called no witnesses. Defendants may have evidence, not presented, that the sole fault in the accident was that of plaintiffs or the owners and operators of the wells. The issue of whether Stockham and Kimray were at fault presents classic factual questions for jury resolution. Fairness to all parties requires a new trial.
For the foregoing reasons, the judgment of the court of appeal is reversed; the judgment of the trial court granting directed verdicts to defendants is vacated; and these consolidated cases are remanded to the trial court for a new trial.
REVERSED AND REMANDED.
NOTES
[1] Reilly v. Dynamic Exploration, Inc., 558 So.2d 1249 (La.App. 1st Cir.1990).
[2] Reilly v. Dynamic Exploration, Inc., 561 So.2d 108 (La.1990).
[3] A check valve is a simple valve, described as a "wear item," a part which is intended to wear out and has to be frequently changed or maintained. The inner workings of the valve can be easily lifted out with a wrench, and the check valve should have had maintenance every six months.
[4] It may be significant that these tests were run with water and Poole's tests were run with air in view of Owen's testimony that check valves are reliable for fluids but not for gas.
[5] The trial judge stated:
"... I'm acquainted with the four categories of what may be unreasonably dangerous. As indicated by the Supreme Court in Halphen, I'm familiar with the elements of proof in a product liability suit." Joint Testimony, Vol. 3, p. 188.