Perkins v. Entergy Corp.

756 So.2d 388 (1999) Don PERKINS, et al. v. ENTERGY CORPORATION, et al. Joseph Bujol, III, et al. Joseph Bujol, III, et al. v. Entergy Corporation, et al. Robert Hracek, et al. v. Entergy Corporation, et al. Nos. 98 CA 2081 to 98 CA 2083. Court of Appeal of Louisiana, First Circuit. December 28, 1999. Opinion Granting Rehearing for Limited Purpose April 13, 2000. *391 Kenneth P. Carter, New Orleans, LA, John A. Braymer, Baton Rouge, LA, Mack E. Barham, Robert E. Arceneaux, Travis B. Bourgeois, New Orleans, LA, for Entergy Corp. and appellants: Entergy Services, Inc., Gulf States Utilities Company and Louisiana Power & Light Company. Victor L. Marcello, Donald T. Carmouche, Gonzales, LA, for appellees, Don A. Perkins, et al. Donald W. Price, Paul H. Due, Baton Rouge, LA, Patrick W. Pendley, Plaquemine, LA, David W. Robertson, Dripping Springs, TX, for appellees, Joseph E. Bujol, III, et al. Edward J. Walters, Jr., Baton Rouge, LA, for appellees, Eric Hracek, et al. Phillip L. Canova, Jr., Plaquemine, LA, for Air Liquide America Corp. *392 Robert H. Wood, Jr., Metairie, LA, for Exxon Corporation. Stephen W. Glusman, Baton Rouge, LA, Lyon H. Garrison, James F. Holmes, New Orleans, LA, for Intervenor Highlands Underwriters Insurance Company. J. Michael Veron, Lake Charles, LA, for Dresser Industries. David F. Bienvenu, New Orleans, LA, Jeffrey A. Goldwater, Chicago, IL, for Westchester Fire Insurance Company. George B. Hall, Jr., Jay Russell Sever, New Orleans, LA, for Agricultural Insurance Company. H. Lee Leonard, LaFayette, LA, for Reliance Insurance Company of Illinois. T. Justin Simpson, Metairie, LA, W. Shelby McKenzie, Baton Rouge, LA, for Lexington Insurance Company and Hartford Casualty Insurance Company. J. Clayton Davie, Jr., Claude F. Bosworth, Loree Peacock Leboeuf, New Orleans, LA, W. Shelby McKenzie, Baton Rouge, LA, for The Aetna Casualty and Surety Company. John P. Wolff, III, Baton Rouge, LA, W. Shelby McKenzie, Baton Rouge, LA, for Federal Insurance Company. Jerry L. Saporito, Michael D. Sledge, New Orleans, LA, W. Shelby McKenzie, Baton Rouge, LA, for Royal Indemnity Company. Julia A. Dietz, Sidney W. Degan, III, New Orleans, LA, D. Russell Holwadel, New Orleans, LA, for National Union Fire Insurance Company of Pittsburgh, PA. Walter C. Thompson, Jr., New Orleans, LA, for X.L. Insurance Company. BEFORE: LeBLANC, GONZALES, GUIDRY, WEIMER, and PETTIGREW, JJ. WEIMER, J. This appeal involves a negligence case brought pursuant to LSA-C.C. art. 2315, et seq. The plaintiffs are two individuals who were severely burned in an industrial accident, Don A, Perkins and Joseph E. "Jeb" Bujol, III, and their families, and the survivors of Ray Hracek, who received fatal burns. The plaintiffs sued Air Liquide America Corporation ("ALAC"), the employer of Mssrs. Bujol, Perkins, and Hracek; Big Three Industries, Inc. ("Big Three") and its insurers; Entergy Services, Inc. ("ESI"), Gulf States Utilities Company ("GSU"), and Louisiana Power and Light, Inc. ("LP & L"), (collectively, "electric utility companies"); Entergy Corporation;[1] Exxon Corporation ("Exxon"); Dresser Industries, Inc. ("Dresser"); and Masoneilan International.[2] A bench trial on the merits was held against the electric utility companies only, the plaintiffs having settled with the other defendants, and the matter was taken under advisement. As stated in written reasons, the trial court determined that the negligence of ESI, LP & L, and GSU was a cause of the accident. The court assessed the electric utility companies with 40 percent of the fault for the plaintiffs' injuries. Additionally, the court determined that ALAC and Big Three were at fault and assessed these defendants with 40 percent and 20 percent of the fault, respectively. Total damages determined by the trial court were $22,728,450.00 (subject to reduction for fault allocations). The electric utility companies appealed. The plaintiffs answered the electric utility companies' appeal; however, we need not address the plaintiffs' answers because of our disposition of this case. Finding that the trial court was manifestly erroneous in holding the electric utility companies[3] liable for plaintiffs' damages, we reverse. *393 BACKGROUND On April 6, 1994, ALAC[4] operated an air separation facility near Plaquemine, Louisiana, which produced oxygen, nitrogen, and argon. Three of the four air separation plants within the facility were producing salable oxygen which was transported via pipeline to customers. An oxygen flash fire and explosion occurred at the ALAC facility approximately three hours after a GSU transmission line located at the Exxon refinery in Baton Rouge, approximately 15 miles away, experienced an electrical fault. The GSU system had recently become part of the Entergy system, which included LP & L and ESI. The plaintiffs' theory is that the electrical fault caused a voltage sag at the ALAC facility which, in turn, loosened debris in the form of metallic particles contained within the ALAC piping system that transported oxygen. Approximately three hours later, after the ALAC facility was restarted, this debris, which had begun to move about within the pipeline, became an ignition agent by what was referred to as "particle impact or frictional induction." The flash fire and explosion occurred at the site of an automatic pressure control valve. FACTS At 1:52 a.m., a static shield wire, held together with only one of its original seven strands and suspended above a 230,000-volt line at the Exxon refinery in Baton Rouge, separated in a 10-knot wind,[5] causing an "A phase to ground" electrical fault. The initial electrical fault was successfully cleared by high speed primary relaying systems in seven cycles,[6] that is, 7/60ths of a second. However, attempts to reclose a breaker caused an escalation to a full three-phase fault. Improperly set and poorly maintained relaying equipment failed, resulting in a 58-cycle, three-phase electrical fault that caused significant loss of voltage on the grid[7] and loss of generation of electricity in the area. The ALAC facility suffered a voltage sag of 47.5 percent. When ALAC experienced the voltage sag, its facility shut down. The nature of the shutdown was the subject of considerable testimony at trial, which we will address later in detail. According to the Schmidt report,[8] three of the four plants had been restarted when a problem was noted in an oxygen pipeline automatic pressure control valve located at the ALAC facility's letdown station. Mr. Perkins was working the shift when the shutdown occurred.[9] Mr. Hracek and Mr. Bujol were among other employees[10] who *394 were called in by ALAC to restart the plants. In an attempt to counteract the problem, the three men went to the letdown station. Shortly thereafter, a large flash fire occurred at the letdown station. Mr. Bujol and Mr. Perkins were blown in different directions by the force of the fireball and suffered burns on approximately 90 percent of their bodies. Mr. Hracek, suffering burns on approximately 95 percent of his body, died five days later.[11] Both lay and expert witnesses testified by deposition and at trial concerning three events on the fateful morning in question: 1) the Entergy transmission system fault at the Exxon refinery, including what preceded and what followed the fault; 2) the shutdown of the ALAC facility; and 3) the startup of the ALAC facility, including the flash fire at the letdown station. 1) Electrical fault at the Exxon Station: Roger L. Owens, vice-president of engineering at McDowell Owens Engineering, a firm specializing in "failure analysis," was called by the plaintiffs and accepted as an expert in electrical engineering and in casualty control. He stated his assignment was to look in detail at the electrical disturbance and then to "look generically... at the explosion fire incident" to see if there was a cause-and-effect relationship. Mr. Owens' description of the electrical fault that caused the voltage sag was basically unrefuted. The first incident was of short duration and consisted of the failure of the static shield wire above the 230,000-volt transmission wires. The static shield wire failed, separated, and fell onto the transmission wires. When contact was made with phase A, the relaying equipment triggered appropriately within seven cycles, which Mr. Owens acknowledged was a "very fast response." Within two minutes, the local operator for the electric utility company attempted reclosure of the line, which is normal procedure. However, the line was reenergized with the downed static shield wire still in contact with the transmission wires. An escalation into a three-phase electrical fault occurred. This time the protective relaying equipment did not perform as designed. The electrical fault continued for 58 cycles which, according to Mr. Owens, was a "huge amount of time in the electrical world." The second incident ended when a manually-recycled breaker was reopened and the fault cleared. Mr. Owens stated the static shield wire fell because the ground strap at the tower, which is a support mechanism, became worn over time. The static shield wire consisted of seven strands, six of which had broken prior to that day. Wind caused the remaining one strand to give way. The static shield wire had been in place 17 to 18 years and was designed to withstand 40- to 45-knot winds, but it gave way in a 10-knot wind. Additionally, the degree of sag in the static shield wire was not within specifications, causing too much tension in the wire. It was Mr. Owens' opinion that the tension was an aggravating factor, not an initiating factor of the failure of the static shield wire. The second electrical fault continued because of a problem with the relay equipment.[12] The noise settings on the relay equipment were set too high, at minus 52 decibels instead of the proper minus 37-decibel setting, and the relay did not sense the electrical fault and signal the breaker. The improper setting was a "huge" difference, in Mr. Owens' opinion. The witness also thought there was no design defect in the relay, but it failed to function properly because it was badly set or not properly reset. *395 The contact settings on the second backup system were misaligned, so that relay did not sense the electrical fault. The third protective device did not function because it was not designed to pick up a three-phase ungrounded fault. Finally, the fourth relay system was an impedancetype, long distance relay that did sense the electrical fault, but only after 58 cycles. Thus, multiple failures in the relay system caused the 58-cycle delay in correcting the electrical fault. This 58-cycle electrical event lasted less than one second, precisely 58/60ths of a second. 2) Shutdown of the ALAC Facility: The electrical fault and failure of the backup systems resulted in a voltage sag. Major equipment at the ALAC facility and the Exxon refinery automatically shut down when the voltage sag was experienced. On cross examination, Mr. Owens stated ALAC never completely lost electrical power. LP & L maintained uninterrupted power during the 58-cycle incident; however, there was a voltage sag which resulted in a voltage drop of 47.5 percent at ALAC. The compressor motors at the ALAC facility shut down even though the electric power was still going into the facility because the voltage sag triggered protective devices designed to prevent large motors from burning up during low voltage operation. If the voltage goes lower than 10 to 12 percent less than specified on the name-plate rating, the protective devices will shut the motors off automatically. Mr. Owens noted, based on his experience, equipment does not shut down in an orderly fashion in an emergency shutdown as contrasted with a manual shutdown. In Mr. Owens' experience, a 58-cycle voltage sag would cause "transients"[13] on the line. Mr. Owens stated the electric utility companies' generator was trying to maintain 230,000 volts; however, in his opinion, the 58-cycle voltage sag created transients consisting of spikes and lags with the voltage way up and then way down. He testified the shutting down of high horsepower motors created more transients. Mr. Owens believed ALAC was experiencing the same transients as on the grid, but stepped down to its voltage level. The witness explained this voltage sag is different from a power shutoff, which has no real shock to the machines and the pipeline flow system.[14] In an effort by plaintiffs to relate Mr. Owens' testimony regarding transients to the plaintiffs' theory concerning the cause of the flash fire, Mr. Owens was questioned about two documents he reviewed: the Schmidt report and the Lowrie article[15]. The Schmidt report describes the plaintiffs' theory of causation, referred to as the "particle impact/frictional induction theory." Mr. Owens praised the Schmidt report as being "meticulously prepared." Concerning the cause of the fire and explosion at the site of the automatic pressure control valve, involving the valve and a straight run of pipe downstream of the valve, the report states: The most likely origin of ignition can be attributed to the following two potential sources: Particle Impact, Friction.... The suspected cause of the primary oxygen fire, originating either in or immediately downstream of the Masoneilan 3" [automatic pressure] control valve, is attributed to foreign material existing within the piping system. *396 [T]he presence of foreign material, i.e., residual metallic particles, within a piping system is not unusual or uncommon even after rigorous cleaning of the piping. Normally, such residual particles are incapable of being transported through the piping system because of their mass or because of insufficient gas velocity and thus remain fix [sic]. However, during highly abnormal pipeline operating conditions, which occurred during the early hours of April 6, 1994, circumstances could have developed which caused the movement of metallic particles through the piping network. Therefore, the potential availability of metallic particles within a piping system and the likelihood of particle impact or frictionally induced ignition becomes a statistical possibility. Mr. Owens stated this particular shutdown was unique,[16] and he pointed out that according to an article authored by Dr. Robert Lowrie, ALAC's expert, startup and shutdown are the most dangerous times in a plant because conditions are not stable.[17] Mr. Owens stated there was no dispute about the electrical upset: it was a long period of time electrically. He had no doubt that transient conditions caused major disturbances at ALAC. When asked his opinion concerning whether there was any cause-and-effect relationship between the voltage sag and the particles becoming loose in the system, Mr. Owens answered: What I'm saying basically in the cause/effect testimony is that based on all of my experiences, which is 35 years in flow systems on ships, submarines, plants, everywhere else, if you go down like that it is very, very likely that you are going to find your strainers and your filters full of debris and particulate matter in systems and things like that. Under those conditions my cause and effect testimony is could that have occurred and then maybe cause the explosion? The answer is, yes. I do not now, my testimony is not it caused the explosion. I'm just saying under those conditions you can break away debris during all this shock and trauma and things like that in the system. What happened there I'm not testifying about. This equivocal statement essentially concluded Mr. Owens' testimony on direct examination. Although Mr. Owens thought the Schmidt report "solidifies" the particulate ignition theory, he did not go so far as to adopt its conclusion. He did not testify as to the specific failure, but was merely familiar with the Schmidt report and the fact that turbulence can cause particulate matter to be broken free. He stated, "I don't think anything was caused after power was restored[;] I think it was all caused in the 58 cycle[s]," but he did not give any reason for that conclusion. However, Mr. Owens admitted he had no knowledge of any specific electrical failures at the ALAC facility. He had no information that the ALAC control systems were compromised by the power disturbance. Although transients can have ill effects on equipment, Mr. Owens had no information that there was an ill effect on any particular hardware at the ALAC facility that morning. Mr. Owens did not look at the specific valve designs to determine exactly what *397 caused the ignition and the resulting detonation within the valve or within the ALAC system. As to the startup procedure, the witness stated ALAC should have notified the power supplier before startup. It was not until Mr. Owens was questioned on re-direct examination that he stated the "most probable" scenario was a "crud burst" or loosening of debris in the ALAC pipeline caused by transients resulting from the 58-cycle voltage sag. Frederick Brooks, accepted as an expert in electrical engineering, electrical utility operating practices, and electrical equipment failures, was called as a witness by the electric utility companies. He concluded the electrical event involving the electric utility companies' lines at the Exxon refinery had no relationship to the flash fire at the ALAC facility because of: 1) the successful shutdown of the ALAC facility, with the motors of the large equipment tripping off at approximately half of the 58-cycle event; 2) the time element between the electrical fault at 1:54 a.m. and the flash fire just before 5:00 a.m.; and 3) the lack of damage to the electrical equipment at the ALAC facility, such as motors, instruments, control systems, all of which would have been susceptible to damage from surges or spikes and/or any other electrical abnormality. As to the presence of transients posited by Mr. Owens, Mr. Brooks testified that on the fault recorder data collected the morning of the event, the voltage sag was limited to a steady drop of 55 percent below normal, with no spikes above 100 percent. Thus, no transients were created, either on the grid or at the ALAC facility. Certainly, no transients were created that could be described as "violent transients that would produce traumatic electrical events." Instead, the voltage sag could be described as a slow down.[18] The only eyewitness to the April 6 shutdown at the ALAC facility who gave live testimony at trial was Mr. Perkins. He stated there were no problems that evening and then "all of a sudden the power was out." He explained that "everything goes black in a split second" and then the facility's generators kicked on. On cross examination, Mr. Perkins explained that each power interruption can be different. Some power interruptions may cause half the plant to shut down, but that morning there was a "total power failure" and everything shut down, when the facility went "from light to dark in the blink of an eye." Ricky J. Webber, the night shift supervisor, gave considerable deposition testimony concerning the startup procedures, but had very little to say about the shutdown. He stated they were sitting around discussing baseball when everything suddenly got dark. Although he stated the shutdown was a total power failure and there was "a lot of noise," he did not elaborate. The deposition testimony of Michael Rockett, the other night shift B operator, did not contain any description of the shutdown. Although Mr. Bujol was at home and thus not present when the plant shut down, he testified concerning past unplanned shutdowns and described them as "real violent." Any time there is an unplanned shutdown of the entire facility, "it's more like a shake, rattle and roll" from the vibrations. Each unit at the plant had a vibration detection system. With excessive vibration, an alarm would sound causing an automatic trip device to shut down the unit. Asked whether he knew if the alarms were activated when the voltage sag occurred on April 6, 1994, he said they would not have been because when the power was cut, the machines already were going down. *398 Both Mr. Hracek, the plant manager, and Anthony J. Mabile, the assistant plant manager, received calls to go to the facility following the shutdown. Mr. Mabile was called by the defense as a fact witness at the trial. He testified the facility experienced four to five power outages per year, and each time the plants would be restarted. Mr. Mabile stated that if the facility loses power for a half second, every compressor shuts down. Perhaps it had been a year since the previous shutdown. When asked what shuts the equipment down, he stated there were electronic relays in the switch gears that monitor the power; when there is a power disturbance, the relays cut the compressors down. Some of the relays protect against under voltage. There is an instantaneous setting on the relays that is set for a particular voltage, for example, for 90 volts out of the normal 110. The disturbance on the morning of the accident caused a shutdown of all of the motors and compressors at the facility. The employees' testimony concerning the shutdown of the facility was noted by the defense witness, Mr. Brooks. The expert's opinion was that a 30-cycle voltage sag, as testified to by Mr. Mabile, would not have had a meaningful effect on the ALAC facility which used synchronous motors.[19] The electrical and mechanical parameters of both the synchronous motors and the reciprocating compressors at the ALAC facility were stabilizing factors that would have reduced the effect of fluctuating voltages. Further, the ALAC facility was not the only plant to experience the voltage sag. All plants between the City of Plaquemine and Exxon in Baton Rouge experienced the voltage sag, but Mr. Brooks knew of no other accidents in the area and there was no evidence of any other accidents. 3) Startup of the ALAC Facility: Mr. Bujol explained the procedure that was necessary to restart the large motors of the particular compressors. He stated all motors have a critical point at which they vibrate more than normal. Thus, there is a "permissive starter" button which can be held down until the motor gets past the high vibration spots. The vibrations will not shut the motor off as long as the permissive starter button is being pressed. Mr. Bujol said the permissive starter buttons had to be used at startup. There was nothing unusual about the startup following this power disturbance. Asked further about vibration, Mr. Bujol stated that oxygen boosters do have some vibration. In the past, prior to the April 6, 1994 incident, vibration had caused sufficient damage to a building at the facility such that ALAC installed braces. On cross examination, Mr. Bujol testified this startup was not the first that he worked after a power disturbance. He had been involved in 10 to 15 startups from the ground up in five years and 30 to 40 partial ones caused by a "blip." The only thing out of the ordinary on the morning of the accident was that the synchronizer pack had failed on the No. 3 plant. The synchronizer pack is "DC power applied and ... keeps the motor from running into lag, which would burn the motor up." Mr. Perkins explained the procedure actually followed after the shutdown. ALAC had customers on line who needed the product produced by ALAC, so it was essential to get the plants back on line promptly. Mr. Perkins testified about the steps taken to restart plants Nos. 1 and 2 within the facility and that there were no problems encountered. Mr. Perkins had participated in 80 to 100 startups in the 10 years he had worked at the facility. Of those, about 20 were *399 whole facility startups like the one on April 6, 1994. When Mr. Mabile got to the ALAC facility, two plants had already been restarted without incident. Plant No. 3 was not up because the synchronizer pack was not functioning. He noted it is possible for the pack to fail during normal operation.[20] At that time, Exxon was down and not in need of product from ALAC, so ALAC was only concerned with supplying its other customers. Mr. Mabile had been involved with dozens of startups after total outages. It takes two to four hours to restart the flow of oxygen. Some time before the accident, Mr. Hracek asked Mr. Mabile to check the system because the pressure to Exxon was low and higher pressure would be needed later. The two men could see the problem from the control room. A device known as a "controller" regulated the automatic valve; it was "telling the valve to go closed and the pressure was still low." However, the controller in the control room, which was set for 720 pounds, was in perfect condition. The next time Mr. Mabile went to the vaporizer system, he passed by the letdown station; he saw the automatic valve wide open. Mr. Mabile had never seen the automatic valve wide open with that pressure before; this condition indicated there was a leak-through problem.[21] The morning of the accident, Mr. Mabile put his hands on the tubing going to the actuator to check it for leaks and found none. About 10 to 15 minutes before the accident, he also tapped on the positioner to determine if the automatic valve would begin to function. However, on cross examination, Mr. Mabile stated he did not think there was an operational problem or a safety problem with the automatic valve on the morning of the accident. Asked about repairs, Mr. Mabile stated there had been no repairs on the automatic valve for several years before the accident. Asked about procedures at the automatic valve, Mr. Mabile stated that typically they would block in the manual valve to solve the problem, which would cut the flow from the 700-pound pipe upstream from the automatic pressure control valve. This manual valve was a back-up to the automatic pressure control valve. During the startup, Mr. Perkins went to the control room and was given instructions to go to the letdown station and close the manual valve. He did not question why.[22] The automatic valve and the manual valve were located a couple of feet apart between the 400- and the 700-pound system. The previous problem with the automatic valve was that it would not close properly; it would allow pressure from the 700-pound pipeline side to escape to the 400-pound pipeline side. Both Mr. Perkins and Mr. Bujol went to close the manual value, as it is large and difficult to turn; it normally took two people and about ten minutes to close it. Mr. Hracek accompanied Mr. Perkins and Mr. Bujol. Mr. Perkins and Mr. Bujol had made a couple of pulls on the handle of the manual valve when Mr. Hracek advised them to stop. They stepped over to look at the automatic valve, and the huge fireball blew. Mr. Perkins testified he could not *400 see if the automatic valve was open or closed because it was too dark in the area near the valve. Mr. Bujol confirmed they went to the letdown station to close the manual valve, which he had closed 15 to 20 times before, because the automatic valve was allowing too much pressure to leak through from the 700- to the 400-pound pipeline. Thus, they could not maintain the pressure on the pipeline supplying Exxon. The leak-through was not supposed to happen. Immediately before the flash fire, Mr. Hracek positioned himself near the automatic pressure control valve. Mr. Bujol told Mr. Hracek he thought the automatic pressure control valve was closed. Then he saw the valve begin to "stroke up."[23] He testified the automatic pressure control valve was opening and then it just dropped, slammed and exploded. Mr. Bujol further testified the valve "can slam shut, I watched it. It may not should slam shut[,] but it did slam shut[;] there is a difference." Mr. Bujol testified he could hear the pressure humming, letting down from 700 to 400 pounds; he could see the workings inside the valve housing. Asked about enough light, Mr. Bujol stated when the valve is open one can see it because there are parts which are "shiny" and thus visible. The automatic valve was all but 5 to 10 percent closed before it began to stoke up. "It stroked up, it never stopped. When it hit the top[,] it just dumped." Mr. Bujol was thrown against a nearby concrete block wall. The six-foot wall was constructed to protect the building. "[A]t other plants [workers were not] even allowed around situations like that." He remembered two explosions or fireballs. The pipeline was like a jet; the ground was rumbling. The electrical disturbance had occurred at 1:54 a.m.; the flash fire occurred at approximately 5:00 a.m. Mr. Bujol knew the automatic valve had stuck before, allowing the pressure to let down from 700 to 400 pounds. Asked about repairs to the automatic valve, Mr. Bujol said there were none. Parts had been ordered and were in the warehouse, but the valve had not been taken down in four years. Asked about concerns for his safety, he said any time one goes out to the letdown station there is concern because of the pressure. Mr. Bujol was especially concerned with opening the manual valve, but he was not concerned with an explosion. There was just talk among the operators that "it was really a pain" to have to adjust the manual valve because the automatic valve leaked. Asked about any problem that morning with pressure from the 700-pound line to the 400-pound line, he stated that if they were going out to the letdown station, there was a reason. He verified that there was an automatic safety release valve. However, there was no way of telling if it was operating unless it opened, and it had not opened. As to debris in the pipeline, the employees' testimony varied. Mr. Perkins said he had no knowledge of crud in the pipeline. Mr. Mabile testified that subsequent to the accident, the heads on each compressor were opened and checked for particles. Nothing was found.[24] Mr. Bujol said the problem with the synchronizer pack had nothing to do with particles in the pipeline. The Lowrie article previously referred to explains in greater detail the part played by particles in the ignition of an oxygen fire and the on-going precautions that must be taken by those in the oxygen industry to prevent such ignition: Design is very important in determining the safety of oxygen equipment. Most metals are difficult to ignite, and design choices can greatly reduce the probability *401 of ignition. Providing adequate clearances and/or rubbing strips of ignition-resistant materials can help prevent fires in rotating equipment. Likewise the use of sensors to detect undue vibration or temperature increases can provide extra time to shut down such equipment before an ignition occurs. Designing for ease of cleaning allows removal of easily ignitable contaminants that might, in turn, ignite a metal part. Providing screens or filters and keeping oxygen gas velocities below critical values can greatly decrease the probability of ignition by particle impact. . . . . The general oxygen compatibility of metals and alloys in various situations can be measured reasonably well by tests of their resistance to promoted combustion and to ignition by rubbing or particle impact.... However, for many applications we must use carbon, alloy, or stainless steels or aluminum alloys.... This is ... done regularly and safely even though the oxygen concentration and pressure are such that these metals can burn. We accomplish this by designing, operating, and maintaining the equipment so as to avoid ignition events to the maximum possible extent and to minimize the damage from an ignition. See n. 15 regarding the Lowrie article. In the "Recommendations" section of the Schmidt report, it was suggested that certain safeguards be implemented to mitigate the presence of foreign material within the ALAC oxygen systems. A principal recommendation was the installation of conical screens upstream of control valves. The Schmidt report also recommended the following, among other suggestions, to reduce the danger to the workers: 1) barrier walls to be utilized around all oxygen control valve stations in high velocity service, such walls designed to withstand forces of oxygen pipeline fires and ruptures; 2) placement of manual valves behind the wall with their operating hand wheels projecting outside the wall for access; and 3) evaluation of the advantages/disadvantages of wearing specialized protective clothing while working around oxygen process equipment. These recommendations are consistent with the trial court's finding that ALAC was negligent in failing to maintain the automatic valve and failing to upgrade its facility when it knew of the dangers of operating the letdown station without adequate filtration devices and without barrier walls with remote controls for the valves, and that ALAC was negligent in allowing its personnel to troubleshoot the automatic valve without depressurizing or isolating it. Also introduced into evidence was the deposition of Dr. Chester Grelecki, a consultant for the chemical industry in the area of fire and explosion hazard evaluation. Attached to his deposition is an article entitled, "Safety Considerations Regarding the Use of High Pressure Oxygen." The article makes it clear that there can be many causes for particulates to move within an oxygen system and thus to ignite.[25] *402 A final witness of note was Sonja Bujol Hracek, a sister of plaintiff Mr. Bujol and a daughter-in-law of the late Mr. Hracek. She testified that Mr. Hracek, who knew he was in critical condition, called her to his bedside and told her he saw the automatic valve slam closed and blow out the bottom. He said, "The damn automatic valve slammed closed on us." She indicated his statement was remarkable in two respects: he never cursed and he was insistent she remember exactly what he said. DISCUSSION Louisiana Civil Code article 2315 states in part: Every act whatever of man that causes damage to another obliges him by whose fault it happened to repair it. Louisiana Civil Code article 2316 states: Every person is responsible for the damage he occasions not merely by his act, but by his negligence, his imprudence, or his want of skill. Articles 2315 and 2316 provide little guidance regarding how to analyze a negligence case. However, the jurisprudence has developed a plethora of approaches for evaluating whether one is negligent and the ultimate issue, whether one should be held liable for negligent conduct.[26] A review of the jurisprudence of Louisiana indicates most cases utilize the duty-risk analysis.[27] This approach provides an analytical framework to evaluate liability. "[T]he duty-risk approach is more policy orientated [than the proximate cause approach] and offers a better opportunity for use of logic and reasoning, particularly when the critical issue in the case involves the scope of protection contemplated by the statute or rule of law which imposed the duty that was breached." *403 Fowler v. Roberts, 556 So.2d 1, 5 (La.1989), reh. granted on other grounds and original opinion reinstated as supplemented, 556 So.2d at 13 (La.1990). Even when the duty-risk analysis is used, there are various approaches that can be applied.[28] Duty-risk is the appropriate analysis in an action brought against an electric utility company for negligence. Lajaunie v. Central Louisiana Electric Company, Inc., 552 So.2d 746, 747 (La.App. 1 Cir.1989), writ denied, 558 So.2d 1130 (1990). Regardless of the approach used, the ultimate issue in a negligence case remains: should this defendant respond in damages to this plaintiff for this action or inaction? One analysis, apparently first proposed jurisprudentially by Justice Lemmon in Fowler, 556 So.2d at 4, requires proof by the plaintiff of five separate elements: (1) the defendant had a duty to conform his conduct to a specific standard (the duty element); (2) the defendant's conduct failed to conform to the appropriate standard (the breach element); (3) the defendant's substandard conduct was a cause in fact of the plaintiff's injuries (the cause-in-fact element); (4) the defendant's substandard conduct was a legal cause of the plaintiff's injuries (the scope of liability or scope of protection element); and (5) the actual damages (the damages element).[29] A negative answer to any of the inquiries of the duty-risk analysis results in a determination of no liability. See Mathieu v. Imperial Toy Corporation, 94-0952, p. 11 (La.11/30/94), 646 So.2d 318, 326. See also, Daye v. General Motors Corporation, 97-1653, p. 9 (La.9/9/98), 720 So.2d 654, 660. We will use the five-element approach[30] outlined above because we conclude it best provides the analytical framework to evaluate whether liability should be imposed in this case. Each element is vigorously contested by the parties. The Duty Element: On appeal, the electric utility companies contend the duty imposed by the trial court, specifically, the duty "to exercise the utmost care to reduce hazards to life as far as practicable,"[31] was not the appropriate duty. A duty has been defined as "an obligation, to which the law will give recognition and effect, to conform to a particular standard of conduct toward another." Morris v. Orleans Parish School Board, 553 So.2d 427, 429 (La.1989). See also W. PAGE KEETON, ET AL., PROSSER AND KEETON ON *404 THE LAW OF TORTS § 53 at 356 (5th ed.1984). The imposition of a legal duty depends on a case-by-case analysis. Gresham v. Davenport, 537 So.2d 1144, 1147 (La.1989); Laiche v. Kohen, 621 So.2d 1162, 1163 (La.App. 1 Cir.1993). Duty is a question of law;[32] the inquiry is whether the plaintiff has any law (statutory, jurisprudential, or arising from general principles of fault) to support the claim that the defendant owed him a duty. Faucheaux v. Terrebonne Consolidated Government, 615 So.2d 289, 292 (La.1993) (citing Green, The Causal Relation Issue and Negligence Law, 60 Mich.L.Rev. 543, 562-63 (1962)). See also Griffin v. Danos and Curole Marine Contractors, Inc., 94-1789, p. 6 (La.App. 1 Cir. 5/5/95), 655 So.2d 525, 528, writ denied, 95-1383 (La.9/15/95), 660 So.2d 451; Ayres v. Beauregard Electric Cooperative, Inc., 94-811, p. 12 (La. App. 3 Cir. 9/6/95), 663 So.2d 127, 134, writs denied, 95-2432, 95-2434 (La.12/15/95), 664 So.2d 455. As a question of law, duty is a legal question subject to de novo review on appeal. See Ferrell v. Fireman's Fund Insurance Company, 94-1252, p. 7 (La.2/20/95), 650 So.2d 742, 747. The duty recited in Levi v. Southwest Louisiana Electric Membership Cooperative (SLEMCO), 542 So.2d 1081, 1084 (La.1989)[33], to exercise utmost care to reduce hazards to life as far as practicable, was not the appropriate duty to impose in this case. Utmost care is the appropriate duty where someone is directly injured due to the dangerous nature of electricity, for example, when someone is electrocuted by coming in contact with a high voltage line in the possession and control of the defendant utility company. Direct contact is not the situation in the instant case.[34] In cases which involve accidents other than electrocution, that occur on the property of a customer and are allegedly caused by some action or inaction on the part of the electric utility company, as in the instant case, the duty is "to use reasonable care in the installation, operation, and maintenance of their electric lines." Boyd v. Washington-St. Tammany Electric *405 Cooperative, 618 So.2d 982, 986 (La. App. 1 Cir.), writ denied, 625 So.2d 1045 (1993). Thus, when the utility company does not own the equipment or lines which directly cause the injury, but merely passes its electricity through the customer's facility, the duty is less onerous than the duty of utmost care. See Northern Assurance Company v. Louisiana Power & Light Company, 580 So.2d 351, 358 (La.1991), (citing Sibley v. Gifford Hill and Company, Inc., 475 So.2d 315 (La. 1985), and Hughes v. Louisiana Power and Light Company, 94 So.2d 532 (La. App. 1 Cir.1957)). In Northern Assurance Company, a case which did not involve electrocution, the court specifically held the trial court erred in imposing the duty of utmost care. In Hughes, the court delineated the duty as follows: "The duty and responsibility of a mere generating company is generally held to be limited to making a proper connection and delivering the electric current to the purchaser's wires and appliances in a manner which, so far as such delivery is concerned, protects life and property, and there is no duty of inspection to see that the purchaser's wires and appliances are in a safe condition and kept so. Accordingly, where wiring or other electrical appliances on private premises are owned and controlled by the owner or occupant of such premises, a company which merely furnishes electricity is not responsible for the insulation or condition of such wiring or appliances and is not liable for injuries caused by their defective condition, to such owner or occupant, .... [However,] knowledge of the defective and dangerous condition of a customer's appliances will charge even a mere generator and supplier of electricity with liability for consequences, where current is thereafter supplied to such defective and dangerous appliances, in which case it is the energizing of the line with knowledge of the conditions, and not the conditions themselves, which forms the basis of liability[.]" Hughes, 94 So.2d at 535 (quoting 29 C.J.S., Electricity § 57, pp. 611-613). Application of the duty of utmost care led the trial court in this matter to err in imposing what, in effect, was a second duty. The trial court stated: The defendants knew or should have known of the risk to plant personnel in restarting a plant after [an] extreme voltage sag event causes a shutdown, especially in the case of an oxygen plant where the dangers of releasing internal metallic particulates during turbulent shutdowns is common knowledge within the industry. This increased safety risk to the plaintiffs resulted directly from the defendants' conduct. The imposition of this duty was error, for this duty imposes an obligation that the electric utility companies are not required to satisfy. As the record reveals, the particulate theory is innate to the oxygen industry, not the electric utility industry. Although the electric utility companies' employees testified they had knowledge of plant operations at the ALAC facility, they spoke from the reference point of electrical needs.[35] Although the employees acknowledged startup and shutdown were problematic, the trial court's imposition of a duty which would effectively require that *406 the electric utility companies have responsibility for knowledge of the myriad of intricate internal operations of each of its innumerable customers is not reasonable, from a practical and an economic standpoint.[36] Electric utility companies are not required to anticipate every possible accident which may occur. Simon v. Southwest Louisiana Electric Membership Corporation, 390 So.2d 1265, 1268 (La.1980). Thus, the trial court erred in imposing on the electric utility companies what amounted to a duty which would include responsibility for the internal operations of this oxygen facility and for guarding against conditions within the facility. In this matter, any duty imposed on the electric utility companies must directly relate to electrical matters, not to internal pipeline operations of this customer.[37] The Breach Element: Although we find that the trial judge erred in applying the incorrect duty of utmost care, we conclude that even with the application of the less stringent duty of reasonable care, the electric utility companies breached that duty. The evidence established this major electrical event was wide ranging and impacted numerous customers. The static shield line was initially strung with too much tension and it broke in a 10-knot wind because six of the seven strands were already broken. Three backup measures designed to prevent the electrical fault from escalating to the extent it did failed to operate. Thus, we conclude the electric utility companies breached their duty of reasonable care.[38] The Cause-in-Fact Element: Cause in fact is a factual question to be determined by the fact finder and, thus, is subject to the manifest error standard of review.[39]See Hardenstein v. Cook Construction, Inc., 96-0829, p. 8 (La. *407 App. 1 Cir. 2/14/97), 691 So.2d 177, 183, writ denied, 97-0686 (4/25/97), 692 So.2d 1093. The plaintiff must prove causation by a preponderance of the evidence. Dabog v. Deris, 625 So.2d 492, 493 (La.1993). Cause in fact usually is a "but for" inquiry which tests whether the accident would not have happened, but for the defendant's substandard conduct. Boykin v. Louisiana Transit Company, Inc., 96-1932, p. 9 (La.3/4/98), 707 So.2d 1225, 1230. Where there are concurrent causes of an accident which nevertheless would not have occurred in the absence of one of the causes, the proper inquiry is whether the conduct under consideration was a substantial factor in bringing about the accident. Id. at n. 10. See also Jones v. Hawkins, 98-1259, 98-1288 (La.3/19/99), 731 So.2d 216, 220. In determining cause in fact, we do not consider whether the conduct was negligent or look to policy considerations, but only consider whether the conduct was a substantial factor leading to the accident, i.e., whether it had a direct relationship to the accident. In making this determination, we consider whether, more probably than not, the conduct complained of was a necessary ingredient of the accident. Miller v. Fields, 570 So.2d 39, 42 (La.App. 4 Cir.1990), writ denied, 575 So.2d 374 (1991), cited in FRANK L. MARAIST & THOMAS C. GALLIGAN, JR., LOUISIANA TORT LAW § 4-1, n. 2 at 83 (1996). The primary evidence presented by plaintiffs on the issue of whether the voltage sag caused the fire and explosion was from the Schmidt report and from Roger Owens, the plaintiffs expert witness. The Schmidt report states that particle impact or friction was the most probable source of ignition. However, the Schmidt report does not causally link, by a preponderance of the evidence, the voltage sag and the release of the particles. The Schmidt report concludes: [D]uring highly abnormal pipeline operating conditions, which occurred during the early hours of April 6, 1994, circumstances could have developed which caused the movement of metallic particles through the piping network [of the ALAC facility]. Therefore, the potential availability of metallic particles within a piping system and the likelihood of particle impact or frictionally induced ignition becomes a statistical possibility. (Emphasis ours.) This conclusion is clearly insufficient to establish cause in fact, which requires proof by a preponderance of the evidence. See Dabog, 625 So.2d at 493. One must prove what probably happened as opposed to what "could" have happened and what is a "statistical possibility."[40] *408 Mr. Owens testified on direct examination concerning the voltage sag: Under those conditions my cause and effect testimony is could that have occurred and then maybe cause the explosion? The answer is, yes. I do not now, my testimony is not it caused the explosion. I'm just saying under those conditions you can break away debris during all this shock and trauma and things like that in the system. What happened there I'm not testifying about. (Emphasis ours.) Mr. Owens' equivocal opinion testimony is insufficient to link the voltage sag and the release of particles to establish cause in fact by a preponderance of evidence.[41] Further, pretermitting whether there was adequate proof of transients in the grid as a result of the 58-cycle voltage sag, there was clearly no proof of transients within the ALAC facility—a necessary component of the plaintiffs' theory that the particles were dislodged by turbulence as a result of the voltage sag. Mr. Owens opined that there were transients within the ALAC facility but offered no objective evidence. When asked by plaintiffs counsel what he based his opinion on, he replied he had "no physical evidence," stating that he reached his conclusion based on a review of documentary evidence and based on his experience. However, he testified he made only a "generic" evaluation and never inspected the equipment within the ALAC facility. Significantly, the evidence was uncontradicted that the ALAC equipment, which included vibration sensors and synchronous motors that limit the impact of voltage disruptions, did not suffer the full impact of the voltage sag. The equipment was designed to respond at a maximum of 30 cycles, which was considerably before the impact of the 58-cycle voltage sag could have been experienced. After reviewing the record, we find no evidence of how the particles loosened and arrived at the site of the flash fire on the morning in question. Further, no one who was at the ALAC facility at the time of the shutdown testified concerning unusual events during this particular shutdown. In fact, Mr. Perkins said everything went out in the blink of an eye. No electrical equipment was found damaged, including equipment sensitive to electrical transients. Only Mr. Bujol testified from personal experience concerning turbulence at the ALAC facility when there was an unplanned, total shutdown of the facility. But Mr. Bujol was not at the facility at the time of this shutdown. Indeed, his testimony proved the opposite of an unusual occurrence[42] on the morning in question *409 because his testimony was that the vibrations occurred every time the facility experienced a total shutdown. His testimony that "it was shake, rattle, and roll" referenced each time there was an unplanned shutdown. Mr. Owens testified he would expect to find debris in the pipeline, but there is no evidence that such debris was discovered.[43] Although the trial court felt the testimony of Mr. Owens was more credible than the testimony of the defendant's expert, Mr. Brooks, objective evidence of record contradicts Mr. Owens' views relative to the ultimate issue of causation. Further, Mr. Owens' testimony on direct examination (quoted previously) is not consistent with his testimony on redirect examination. A credibility evaluation by the trier of fact is entitled to great deference pursuant to the manifest error standard of review, but the concept does not mean that the reviewing court virtually can never upset the trial court's findings. Barry v. Plaquemine Towing Corporation, 96-0979, p. 3 (La.App. 1 Cir. 8/4/97), 698 So.2d 1017, 1019, writ denied 97-2579 (1/9/98), 705 So.2d 1102. After giving great deference to the fact finder, this court has every right to determine whether the trial court's finding was clearly wrong based on the evidence or clearly without evidentiary support. Id. We believe Mr. Owens' testimony was credible regarding the failure of the static shield wire and the backup electrical fault correcting devices, as well as his characterization of the 58-cycle fault as a significant electrical event. However, we find his admittedly "generic" evaluation of the impact of this electrical event on the ALAC facility fails to establish the voltage sag was a substantial cause of the flash fire and explosion which occurred three hours later. Thus, although there was sufficient evidence to establish this voltage sag was a significant electrical event, the record does not support a finding the voltage sag was a significant electrical event within the ALAC facility. Based on the evidence previously discussed, we conclude the record does not include a reasonable factual basis for the trial court's finding that the electric utility companies' negligence was a cause in fact of the plaintiffs' injuries. The trial court was clearly wrong in determining the plaintiffs satisfied their cause-in-fact burden of proof. Although we find there was insufficient evidence of a connection between the voltage sag and the flash fire, we choose not to decide this case on only the cause-in-fact inquiry. We will also review this matter based on the scope-of-duty element (legal cause), an issue which we believe predominates over every other issue in this case. The Scope-of-Duty Element (Legal Cause): Whether a duty is owed by one party to another is a legal question; but, "[t]here is no `rule' for determining the scope of the duty." Roberts v. Benoit, 605 So.2d 1032, 1044 (La.1991). The scope-of-duty inquiry is fact sensitive and ultimately turns on "a question of policy as to whether the particular risk falls within the scope of the duty." Id.[44] *410 As previously indicated, the electric utility companies owed a duty to use reasonable care in the installation, operation, and maintenance of their electrical lines. Whether this legal duty encompassed the risk visited upon these particular plaintiffs is the larger question. Recently, in Todd v. State, Department of Social Services, Office of Community Services, 96-3090, pp. 6-7 (La.9/9/97), 699 So.2d 35, 38-39, the Louisiana Supreme Court, stating unequivocally that legal cause is purely a legal issue,[45] explained: "Cause" in legal cause demands an inquiry into whether a legal standard of care exists and requires delving into policies for and against extending the asserted legal standard of care to protect the particular plaintiff against the particular harm. See Malone, Ruminations on Cause-In-Fact, 9 Stan.L.Rev. 60 (1956); Robertson, Reason Versus Rule In Louisiana Tort Law: Dialogues on Hill v. Lundin & Associates, Inc., 34 La.L.Rev. 1 (1973). Moreover, whereas the question of cause-in-fact involves a factual determination, the determination of legal cause involves a purely legal question. See Dixie Drive It Yourself Sys. v. American Beverage Co., 242 La. 471, 137 So.2d 298 (1962). (Emphasis ours.) Every negligence case must be decided on its own facts and circumstances. Roberts v. Benoit, 605 So.2d 1032 (La. 1991). In some instances a risk may not be found within the scope of a duty where the circumstances of that particular injury to that plaintiff could not be reasonably foreseen or anticipated, because there was no ease of association between that risk and the legal duty. See Hill v. Lundin & Associates, Inc., 260 La. 542, 256 So.2d 620 (1972). We are mindful that foreseeability, as the determining test, is neither always reliable nor the only criterion for comparing the relationship between a duty and a risk. Some risks that arise because of a defendant's conduct are not within the scope of the duty owed to a particular plaintiff simply because they are unforeseeable. The ease of association of the injury with the rule of conduct that is urged, however, is the proper inquiry. Hill, supra. Nevertheless, the extent of protection owed a particular plaintiff is determined on a case-to-case basis to avoid making a defendant an insurer of all persons against all harms. Malone, supra. Thus, legal cause can be evaluated on the basis of foreseeability and ease of association. For the purpose of a legal cause determination, we can assume a breach of duty which is the cause in fact of the plaintiff's damages. See Robertson, 58 La. L.Rev. at 17, n. 70. Foreseeability: In this matter, Mssrs. Hracek, Perkins, and Bujol, with their combined 42 years of experience working in the oxygen producing industry[46] and with their involvement in innumerable shutdowns[47] did not foresee the voltage sag leading to the explosion and fire.[48] We are not in any way suggesting Mssrs. Hracek, Perkins, and Bujol were contributorily negligent; rather, it is axiomatic that Mr. Hracek, as the facility manager, would not place himself and his fellow employees in a position of *411 peril if this explosion and fire had been foreseeable. If it was not foreseeable to these persons with experience within the oxygen producing facility that this explosion and fire would occur in this fashion, it was clearly not foreseeable to the electric utility companies, outsiders to the oxygen producing industry, that this tragic event would occur. Ease of Association: The legal cause issue in Sibley, 475 So.2d 315, is very similar to the issue at hand.[49] In that case, the boom of a crane on the premises of the utility company's industrial customer came so close to an overhead power line that electrical arcing occurred, causing fatal injury to one man and serious injury to another. The plaintiffs contended that the arcing was caused by an electrical power surge. The Louisiana Supreme Court recognized that the outcome of the case depended on the extent to which the court would impute liability to an electric company for unusual accidents occurring within the plants of their customers. The court stated: While the causal relation issue involves an analysis of the defendant's conduct as a contributing factor in the injury, the duty issue is a policy inquiry into whether the defendant's duty to the victim included protection against the particular injury. Green, The Causal Relation Issue in Negligence Law, 60 Mich.L.Rev. 543 (1962). LP & L arguably owed a duty to its customers to prevent surging because the customer's plant or equipment might be damaged by a sudden voltage surge. However, this duty did not reasonably encompass the risk that a grounded person on the customer's premises might place a metal object close enough to the transmission line for arcing to occur contemporaneously with the surging, yet far enough away to avoid close proximity arcing. Moreover, there is no ease of association between the duty breached and the damages sustained. Hill v. Lundin & Associates, Inc., 260 La. 542, 256 So.2d 620 (1972). (Emphasis in original.) We conclude that any voltage surge created by LP & L's equipment was not a legal cause of the injuries.... Sibley, 475 So.2d at 319. The court further stated there was no support in the record for a conclusion that LP & L supplied electricity over a private transmission line which it knew or should have known was unreasonably dangerous. Sibley, 475 So.2d at 319-320. When we consider the facts of the instant case in light of the principles set forth in Sibley, the ease of association is clearly more attenuated here, where the particle or friction induced explosion occurred three hours after the voltage sag, than in Sibley, where an electrocution occurred simultaneously with the power surge. When considering limitation of liability issues, whether phrased in terms of scope of duty or legal cause or proximate cause, it can be useful to consider whether "too much else has intervened—time, space, people, and bizarreness." See Roberts, 605 *412 So.2d at 1058 (quoting David W. Robertson, Reason Versus Rule in Louisiana Tort Law: Dialogues on Hill v. Lundin & Associates, Inc., 34 La.L.Rev. 1 (1973)). In this matter, a little over three hours elapsed between the voltage sag and the explosion (time), the electrical fault occurred 15 miles away from ALAC (space), ALAC's actions played a substantially more direct role in the explosion (people)[50], and the Schmidt report references "event randomness" (bizarreness).[51] Thus, liability should not be imposed. In sum, the voltage sag was not a legal cause of the injuries received in the non-electrical, random-event flash fire in the ALAC facility. It is manifestly erroneous[52] to conclude either foreseeability *413 or ease of association was established in this matter.[53] The duty of the electric utility companies to maintain their equipment in a reasonable manner so as to avoid an interruption of electrical service does not encompass the risk that a voltage sag might cause foreign particles in an improperly maintained pipeline of a customer to loosen, then ignite three hours later, after two oxygen plants were restarted without incident, at a time when the customer's employees were examining a malfunctioning automatic pressure control valve. If any of a number of situations had not existed, or circumstances had not occurred, the plaintiffs would not have been injured. The only relationship the electrical outage had to the explosion was the outage required the plant to be restarted, something that had occurred innumerable times without incident. CONCLUSION Because of the horrible injuries suffered by the plaintiffs, we have struggled mightily with the issues presented in this matter. It was impossible not to be emotionally moved when reading the testimony of the plaintiffs and the testimony of their family members, co-workers, and physicians. However, as we instruct juries, our decision must be made, not based on sympathy, but based on the facts and law. Our resolution of the issues in no way reflects on the efforts of the attorneys for the plaintiffs, who represented their clients competently and conscientiously. All parties submitted exemplary briefs. We reverse the judgment of the trial court awarding damages to the plaintiffs. On appeal, parties are to bear their own costs. REVERSED AND RENDERED. PETTIGREW, J., dissents and assigns reasons. GUIDRY, J., dissents and assigns reasons. PETTIGREW, J., dissenting. I must respectfully dissent from the majority. I do find an ease of association between the voltage sag and the ultimate explosion in this case. Hill v. Lundin and Associates, Inc., 260 La. 542, 256 So.2d 620 (1972). Therefore, the breach of the duty owed by defendant was one of the legal causes for the damages sustained by plaintiffs. However, I do feel the trial court was manifestly erroneous in assessing and allocating appellants with 40 percent of the fault. I feel the highest reasonable percentage of fault that the appellants, Louisiana Power & Light Company, Gulf States Utility Company, and Entergy Services, Inc., should be assessed under the particular facts of this case is 10 percent. GUIDRY, J., dissenting. Because I am of the opinion that the trial court's decision was not manifestly erroneous or clearly wrong, I respectfully dissent from the majority opinion for the following reasons. *414 The trier of fact has great discretion regarding factual matters, Creekmore v. Elco Maintenance, 94-1571, p. 3 (La.App. 1st Cir.6/30/95), 659 So.2d 815, 817, and the trial court's findings are entitled to great weight and should not be disturbed where there is, as here, a reasonable basis for its findings. U.S. Oil of Louisiana, Ltd. v. Louisiana Power and Light Co., 350 So.2d 907, 911 (La.App. 1st Cir.), writ denied, 351 So.2d 775 (La.1977). Thus, if the factfinder's findings are reasonable in light of the record reviewed in its entirety, the court of appeal may not reverse, even if convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Banks v. Industrial Roofing & Sheet Metal Works, Inc., 96-2840, p. 7 (La.7/1/97), 696 So.2d 551, 556. A court of appeal may not set aside a trial court's finding of fact in the absence of manifest error. Rosell v. ESCO, 549 So.2d 840, 844 (La.1989). The issue to be resolved by a reviewing court is not whether the trier of fact was right or wrong, but whether the factfinder's conclusion was a reasonable one absent manifest error. Stobart v. State through Department of Transportation and Development, 617 So.2d 880, 882 (La.1993). The idea behind this principle of review is founded upon the trial court's capacity to evaluate live witnesses, as compared with the appellate court's access only to a cold record. The principle also stems from the proper allocation of trial and appellate functions between the respective courts. Thus, where two permissible views of the evidence exist, the factfinder's choice between them cannot be manifestly erroneous or clearly wrong. Ranger Insurance Company v. Mancuso, 94-464, p. 5 (La.App. 5th Cir.2/15/95), 652 So.2d 28, 31. The majority using "manifest error" language throughout its opinion, however, in my opinion, conducted a de novo review and substituted its judgment for that of the trial court. In conducting its review of whether the plaintiffs proved, by a preponderance of the evidence, all the required elements of negligence to prevail in this case, the majority conceded that the electric utility companies had a duty to exercise reasonable care and that they breached this duty.[1] Thereafter, the majority concluded that this breach was not a cause in fact or legal cause of the plaintiffs' injuries. The majority's presentation of the case, however, does not, in my opinion, negate the evidence in the record that supports the conclusion of the trial court. I, after thoroughly examining the evidence in the record, carefully reviewing the trial court's "Reasons for Judgment," and giving the trial judge his due deference, am of the opinion that there is ample evidence in the record to provide a reasonable basis for the trial court's finding of negligence on the part of the electric utility companies. The record in this case, viewed in its entirety, reflects that two permissible sides of the story were presented at trial. A correct application of the "manifest error" standard requires that this court not reverse the trial court's acceptance of one permissible view of the evidence over another. Because the majority, in my opinion, misapplied this standard, I dissent. *415 Before: LeBLANC, GONZALES, GUIDRY, WEIMER, and PETTIGREW, JJ. WEIMER, J., on rehearing. We granted a rehearing for the limited purpose of addressing allotment and rehearing/reargument issues raised by the plaintiffs in their rehearing applications. This matter was initially assigned to a three-judge panel consisting of Judges Gonzales, Guidry, and Weimer. Because of the amount of the award, the voluminous record, and the complexity of the numerous issues—issues which potentially went beyond duty, breach, cause in fact and legal cause, and included whether the independent corporate entities could be solidarily liable, the propriety of the trial court's allocation of percentages of fault, quantum, and calculation of fault with credits for settling tortfeasors—the panel conferred and unanimously decided to have this matter initially decided by a five-judge panel.[1] It was deemed necessary to have this case heard initially by a five-judge panel so as to promote justice and expedite the business of the court. See Uniform Rules—Courts of Appeal, Rule 1-5. It was hoped that potential delays inherent in having the matter initially heard by a three-judge panel could be avoided so that the parties could be afforded an expeditious resolution of the case. The three judges of the panel proceeded to the clerk of court's office and, based on the process customarily employed by the court, two additional judges were randomly selected in the presence of the entire original panel to complete the five-judge panel. The parties were notified and no complaints were voiced prior to oral argument regarding the fact the case was to be heard initially by a five-judge panel. Following the rendition of the judgment by the five-judge panel, the plaintiffs requested a rehearing. Citing La. Const. art. V, § 8, the plaintiffs/applicants argue that "[s]ince two judges dissented after the initial argument of the case, the constitution requires reargument, and thus applicants are constitutionally entitled to a rehearing. The case was only argued once, and the constitution mandates reargument when only one judge dissents." They also cite Uniform Rules—Courts of Appeal, Rule 1-5. Additionally, a contention is made in the rehearing application that the case should be reconsidered en banc. Louisiana Constitution article V, section 8 provides in pertinent part: § 8. Courts of Appeal; Circuits; Panels; Judgments; Terms Section 8. (A) Circuits; Panels. The state shall be divided into at least four circuits, with one court of appeal in each. Each court shall sit in panels of at least three judges selected according to rules adopted by the court. (B) Judgments. A majority of the judges sitting in a case must concur to render judgment. However, in civil matters only, when a judgment of a district court is to be modified or reversed and one judge dissents, the case shall be reargued before a panel of at least five judges prior to rendition of judgment, and a majority must concur to render judgment. The Uniform Rules—Courts of Appeal, Rule 1-5 provides: Rule 1-5. Panels The court ordinarily will sit in rotating panels, each composed of 3 Judges, as may be directed by the Chief Judge. *416 In civil cases, when a judgment or ruling of a trial court is to be modified or reversed and one judge dissents, the case shall be reargued or resubmitted before a panel of at least 5 Judges, a majority of whom shall render judgment. When an appeal is taken from an election case objecting to candidacy or contesting an election, the case shall be heard by the court as directed by law. When authorized by law, or when the court deems it necessary to promote justice or expedite the business of court, the court may sit in panels of more than 3 judges or en banc. Clearly, these provisions contemplate that if a case is heard initially by a three-judge panel and one judge dissents from the modification of the decision of the district court, then the matter is to be heard by a five-judge panel. The Constitution and the appellate court rule do not require that a matter be heard initially by a three-judge panel. This court, inherently and by rule, has authority to initially hear matters by a five-judge panel. The Constitution, which states "each court [of appeal] shall sit in panels of at least three judges" clearly authorizes more than three judges to sit on a panel. (Emphasis added.) La. Const. art. V, § 8(A). Uniform Rules—Courts of Appeal, Rule 1-5 specifically provides that a case may be initially heard by more than three judges, as follows: "When authorized by law, or when the court deems it necessary to promote justice or expedite the business of court, the court may sit in panels of more than 3 judges or en banc." Plaintiffs claim entitlement to a rehearing. However, it should be noted that neither La. Const. art. V, § 8(B) nor Uniform Rules, Rule 1-5 provides for a rehearing; both provisions provide for reargument. Rehearing and reargument are different. A rehearing is requested after a judgment is rendered and a party wants the judgment modified.[2] Reargument is additional argument before the judgment is rendered.[3] When a case is heard initially by a three-judge panel, a party has a constitutionally provided right to reargument because a judgment cannot be rendered in a civil matter when one of the original three-judge panel members dissents and agrees with the decision of the lower court. La. Const. art. V, § 8(B). However, once a judgment is rendered by a five-judge panel, there is no right to reargument and no right to a rehearing; a rehearing is discretionary. In this matter, the plaintiffs did not request reargument before the judgment was rendered. Although the plaintiffs did submit a request to supplement their briefs following oral argument,[4] which request noted the case had been argued only once before a five-judge panel, the issue of a constitutional entitlement to reargument was not advanced. While not suggesting a waiver of any right to reargue, because such a right does not exist when the matter is heard initially by a five-judge panel, certainly the failure to timely raise an issue prevents this court from addressing the issue. It is too late after an unfavorable judgment has been rendered against him for the plaintiff to *417 claim a right that he could have claimed, but failed to claim, before an unfavorable judgment. Duffy v. Throwbridge, 335 So.2d 30, 32 (La.1976).[5] The plaintiffs cite Bank of New Orleans and Trust Company v. Seavey, 383 So.2d 354 (La.1980), for the proposition that judgment should not have been rendered by this court absent a reargument. The holding in Seavey is simply inapposite in the instant case. Seavey involved the issue of whether a five-judge panel was constitutionally required when the court of appeal was exercising its supervisory jurisdiction. The case did not address the issue of a constitutional entitlement to reargument after a case is heard initially by a five-judge panel. Both Seavey and Duffy involved matters heard by three-judge panels initially. Neither case concerned the present facts which involve consideration of the case by a five-judge panel initially. Likewise, the opinion in Sarpy v. Sarpy, 359 So.2d 750, 751 (La.App. 4 Cir.), writ denied, 360 So.2d 1178 (1978), contained language concerning reargument before a larger panel, but the factual setting was that of an initial argument before a three-judge panel. The plaintiffs further contend that the case should be heard before the court sitting en banc. However, there is no provision in the Constitution or the Uniform Rules—Courts of Appeal that entitles one to an en banc consideration after a judgment has been rendered. The Constitution requires that a majority of the judges sitting in a case must concur to render judgment. La. Const. art. V, § 8(B). The purpose of having a case heard initially by a five-judge panel is to assemble a panel of judges so the business of the court can be expedited and a majority can render a judgment without the delays inherent in reargument. Once the matter is heard by a five-judge panel, and a judgment is rendered, there is no entitlement to reargument or en banc consideration. Uniform Rules, Rule 1-5, mentions discretionary en banc consideration. To have the matter argued before a five-judge panel and then reargued before the court en banc would defeat the purpose of having the matter initially heard by the five-judge panel and would result in unnecessary delay. REHEARING GRANTED WITHOUT ORAL ARGUMENT FOR THE LIMITED PURPOSE OF ADDRESSING THE ALLOTMENT AND REHEARING/REARGUMENT LEGAL ISSUES; APPLICATION FOR REARGUMENT AND EN BANC CONSIDERATION DENIED; REHEARING DENIED AS TO THE MERITS. PETTIGREW, J., concurs in part and dissents in part. GUIDRY, J., dissents. PETTIGREW, J., concurring in part and dissenting in part. I agree with the majority as to the denial of an En Bank hearing, the allotment and reargument issues, however I dissent with the majority in that I would grant a rehearing on the merits to the parties before the original 5 Judge panel. GUIDRY, J., dissenting. I disagree with the majority in that I would grant re-argument and en banc consideration on the merits and in the alternative, I would grant rehearing with oral argument on the merits before the original 5 Judge Panel. NOTES [1] Entergy Corporation, the parent corporation of the three electric utility companies that were cast in judgment, was not cast in judgment and, thus, is not an appellant. [2] Dresser and Masoneilan International manufactured the automatic pressure control valve (also referred to as the automatic valve) located at the letdown station, which was the site of the flash fire. [3] ESI, GSU, and LP & L collectively assign as error the trial court's decision that because they had "intertwined roles" their combined negligence was a legal cause of the accident. It is unnecessary to address this assignment of error. [4] ALAC, the employer of Mssrs. Hracek, Perkins, and Bujol, was a corporate successor to Big Three and took over the facility in 1986. [5] The court's reasons for judgment describe a 10-mile-per-hour wind, but testimony at trial indicates it was a 10-knot wind. The difference is inconsequential. [6] An alternating electrical current cycle is a wave form of electrical current from zero to one hundred above or positive, back to zero, then to one hundred below or negative, and finally, back to zero. A cycle takes 1/60th of a second. [7] A grid is the utility alignment with multiple sources powering it to maintain electrical service to a host of customers simultaneously, instead of individual lines to various customers. A huge generator transmits millions of volt amps to the grid. [8] This document, authored by L.J. Schmidt and introduced into evidence, is a report of a committee assembled to investigate the oxygen flash fire and pipeline rupture at the ALAC facility on April 6, 1994. [9] There were four men at the facility, including Mr. Perkins, who was a B operator, when the power outage occurred. Ricky Weber was the shift supervisor, Richie Landry was an A operator, and Mike Rockett was a second B operator. [10] In addition to Mr. Bujol, who was a B operator and Mr. Hracek, who was a manager, five men were called to the facility and arrived prior to the accident. They were: Lawrence Deville, an A operator; Ray Frouix, electrician; Tony Mabile, assistant manager; Kenny Rivet, instrumentation leader; and Larry Prejean, electrician. [11] Dr. Edward Chiasson testified that Mr. Bujol and Mr. Perkins were two of the most severely burned individuals he had ever encountered who were able to survive. Their survival is a tribute to them and their families. Our condolences are extended to the family of Ray Hracek. [12] A relay is a sensing device which tells the breakers when to open. Essentially, the breaker is a large switch. [13] Other testimony in the record defines "transients" as any departure from the norm in electrical cycles. [14] Throughout the trial, various witnesses referred to three types of shutdowns: a planned shutdown, which is orderly because the machinery is manually turned off; a shutdown because of a power outage; and a shutdown because of a voltage sag. Obviously, the planned shutdown is irrelevant in this factual situation. The parties debate the differences between the power outage shutdown and the voltage sag shutdown and which type occurred prior to the explosion. [15] This document, also introduced into evidence, is a paper entitled "Oxygen Compatibility of Metals and Alloys" and written by Dr. Robert Lowrie. The focus of the article is safety in the material selection, design, operation, and maintenance of oxygen equipment to avoid ignition events and to minimize damage from an ignition. [16] ALAC had lost power on several different occasions prior to this date, but Mr. Owens found nothing in the materials he reviewed indicated a magnitude of both voltage and time to equal the April 6 disturbance. Mr. Owens opined ALAC was not required to protect against a 58-cycle event, which is not a normal disturbance. The customer should not have to protect against a 58-cycle event with transients and a 50 percent reduction in voltage. On re-direct examination, Mr. Owens admitted it may have been a 45.7 percent dip in voltage instead of a 50 percent dip. [17] This sentence in the Lowrie article is followed by the following cautionary language: Resonance peaks with the attendant danger of excessive vibration and rubbing can occur in rotating machines. High gas velocities with danger of particle impact ignition as well as adiabatic compression at dead ends can occur if care is not taken. A compressor is often started on nitrogen to check it out before oxygen is admitted. Metal wear dust formed during nitrogen operation will not be oxidized and may be pyrophoric. The change to oxygen should be gradual to slowly oxidize such material. [18] We note that this portion of the witness's testimony was not objected to by counsel for the plaintiffs. An objection was entered shortly after this testimony when counsel for the defendants attempted to have the witness review the fault recorder data. Plaintiffs' counsel objected because the document reflecting the fault recorder data had not been provided during discovery. The objection was sustained and the documentary evidence was proffered. We do not consider whether the proffered material was admissible because the exclusion of the data was not assigned as error. [19] Synchronous motors serve as drivers of electricity. As long as they are spinning, they act as generators and pump power into the system. They help hold up the voltage and stabilize the system. [20] Mr. Brooks verified this testimony, explaining that the synchronizer pack can fail during service, but the failure is only discovered when there is a shutdown and startup. There was no evidence that the synchronizer pack failed during the shutdown or as a result of the voltage sag. [21] Concerning the letdown station, Mr. Mabile stated that both the manual and automatic valves were leaking. ALAC took the system out of service after the accident. Some years before the accident ALAC knew one manual valve was leaking. [22] When questioned concerning "numerous" problems with the automatic valve, Mr. Perkins stated it depended on the person you asked. If he had to close the manual valve, it was because the automatic valve was not working the way it should. That condition had gone on for a while. He had been out to close the manual valve 20 to 25 times before. [23] Although the transcript reads "stroke," the proper word may be "stoke." [24] Mr. Mabile testified that at one compressor some welding slag was found at the in-going suction, none at the outlet side, but this slag was found a few months after the accident. [25] The article cautions: Care should be exercised in installing system piping, e.g., keep the number of bends in the piping to a minimum, since every bend is a potential location for a particle to impact with spark-producing force. Flow control components such as valves ... should be kept to the minimum number that system performance requirements permit. Each component must be considered as having several potential sources of ignition, such as possible generation of particulate during operation, the soft goods in the component, mechanical or pneumatic impact phenomena associated with component operation (for example, forceful contact of a valve stem closing against its seat, or adiabatic compression occurring downstream of a rapidly opening valve), localized heating due to friction on moving parts, entrapped contamination which may not have been removed by the cleaning process, use of improper types or excessive quantities of lubricants on interior moving parts, and so on.... The faster the valve cycle, the greater the chance of ignition due to adiabatic compression, high friction-induced temperatures, etc. . . . . To assure that the system is maintained free from particulate contamination, the oxygen storage and distribution system should be equipped with sufficient filtration to halt the migration of particles into critical components or subsystems. As a minimum, filtration should be provided on the inlet port of all oxygen storage vessels, on the inlet to each subsystem and downstream of any process or component which is a known particle generator.... Minimize personnel exposure to systems storing or flowing high pressure oxygen, especially while components are being actuated or processes involving potential reactions are underway. [26] See Thomas C. Galligan, Jr., A Primer on the Patterns of Negligence, 53 La.L.Rev. 1509, 1510 (1993), which discusses four prominent approaches or "formulas": the standard common law proximate cause approach, violation of statute, duty-risk, and the Learned Hand negligence formula. See also David W. Robertson, Allocating Authority Among Institutional Decision Makers in Louisiana State-Court Negligence and Strict Liability Cases, 57 La.L.Rev. 1079 (1997), which discusses six approaches for analyzing negligence. It has been suggested it would be wise and prudent to adopt one approach. See Robertson, 57 La.L.Rev. at 1102: "We can have a reasoned debate about which approach to select, but there is very little to be said for vacillating among the [approaches] to the extent that lawyers are significantly uncertain as to how to try their cases." See also Robertson, 57 La.L.Rev. at 1104 for a discussion of the vacillation in the jurisprudence between various approaches. [27] The duty-risk analysis is the standard negligence analysis employed in determining whether to impose liability under LSA-C.C. art. 2315. Mathieu v. Imperial Toy Corporation, 94-0952, p. 4 (La.11/30/94), 646 So.2d 318, 321. But see discussion in FRANK L. MARAIST & THOMAS C. GALLIGAN, JR., LOUISIANA TORT LAW § 3-4, n. 14 at 79-80 (1996), regarding Pitre v. Opelousas General Hospital, 530 So.2d 1151 (La.1988), which used a "legal cause" analysis. The authors also discuss the approach of Justice Harry T. Lemmon in Fowler v. Roberts, 556 So.2d 1, 4-5 (La.1989), reh. granted on other grounds and original opinion reinstated as supplemented, 556 So.2d 1, 3 (La.1990), stating: "Although Justice Lemmon referred to his approach as duty/ risk, there is an obvious correlation between the Lemmon approach, the common law approach, and Pitre's "legal cause" approach." MARAIST & GALLIGAN, § 3-4, n. 14 at 80. See also MARAIST & GALLIGAN, §§ 5-4 & 5-5 at 103-108. [28] Some cases refer to a four-element duty-risk analysis. See Mart v. Hill, 505 So.2d 1120, 1122 (La.1987) and Faucheaux v. Terrebonne Consolidated Government, 615 So.2d 289, 292 (La.1993). Other cases refer to a five-element duty-risk analysis. See Fowler, 556 So.2d at 4 and Wilson v. State, Department of Public Safety and Corrections, 576 So.2d 490, 493 (La.1991). See also David W. Robertson, The Vocabulary of Negligence Law: Continuing Causation Confusion, 58 La.L.Rev. 1, n. 12 at 4, which discusses various five-element approaches. Still other cases reference both the four-element analysis and the five-element analysis. See Mathieu, 94-0952 at 4, 646 So.2d at 321-322 and Rhodes v. State, Department of Transportation and Development, 94-1758R (La.App. 1 Cir. 12/20/96), 684 So.2d 1134, writ not considered, 97-0242 (2/7/97), 688 So.2d 487. Some cases list three elements of a cause of action based on LSA-C.C. arts. 2315 and 2316 as "fault, causation, and damage" while also referencing "duty" and "risk." See Dabog v. Deris, 625 So.2d 492, 493 (La.1993) and Morris v. Orleans Parish School Board, 553 So.2d 427, 429 (La.1989). [29] "The first element is usually a judge question, and the other four are usually jury questions unless reasonable minds could not differ." Fowler, 556 So.2d at 4-5. This statement is dicta; the case did not involve a jury trial. [30] See David W. Robertson, The Vocabulary of Negligence Law: Continuing Causation Confusion, 58 La.L.Rev. 1, n. 12 at 4, suggesting this five-element approach is the most common and clearest of the various approaches. [31] The trial court cited Levi v. Southwest Louisiana Electric Membership Cooperative (SLEMCO), 542 So.2d 1081 (La.1989) as authority for the imposition of this duty. [32] See MARAIST & GALLIGAN, § 3-2 at 76: "Generally, duty is considered a question of law for the court." Also see § 5-3 at 102 (1996). [33] Significantly, the Louisiana cases cited by the trial court along with Levi were all electrocution cases. The duty of utmost care in electrocution cases is based on the rationale that where the wires maintained by an electric utility company are designed to carry a powerful current of electricity, so that persons coming in contact with them are certain to be seriously injured, if not killed, the law imposes upon the utility company the duty of exercising the utmost care and prudence consistent with the practical operation of its plant to prevent such injury. The electric utility company must provide such protection as will safely guard against any contingency that is reasonably to be anticipated; however, it is not legally bound to safeguard against occurrences that cannot be reasonably expected or contemplated. Calton v. Louisiana Power & Light Co., 56 So.2d 862, 865 (La.App. 2 Cir. 1952), aff'd, 222 La. 1063, 64 So.2d 432 (1953). The courts have further defined the duty in cases where the injury, such as electrocution, occurs with the escape of electricity while being transmitted over high power lines, as threefold: (1) to insulate the lines, or (2) to warn adequately of the danger, or (3) to take other proper and reasonable precaution to prevent injury. Simon v. Southwest Louisiana Electric Membership Corporation, 390 So.2d 1265, 1267 (La.1980). [34] In discussing duty, the trial court cited one Louisiana case that did not involve an electrocution, U.S. Oil of Louisiana, Ltd. v. Louisiana Power & Light Co., 350 So.2d 907 (La.App. 1 Cir.), writs denied, 351 So.2d 775(1977). The trial court stated this court affirmed a damage award against LP & L for losses sustained by a plant resulting from power outages, including damages for the additional costs incurred in restoring the plant to normal operations and for loss of plant production. The cited case involved three power outages on three separate dates. This court found LP & L was not liable for the damages following two of the outages. The damage awards were made after LP & L stipulated to being liable for the damages following the third outage. Thus, the precedential value of U.S. Oil of Louisiana, Ltd. in the instant case is, at best, questionable. [35] The trial court's determination regarding knowledge of the internal operations of the ALAC facility is not a fair assessment of the testimony of the electric utility companies' employees. They clearly testified they had limited knowledge of the internal workings of the oxygen facility. The depositions of three representatives of the electric utility companies, George M. Lee, Jr., Joseph A. Gavin, and Timothy E. Ponseti, a senior account representative with ESI, were introduced into evidence. All three men denied having any knowledge of the Internal operational aspects of the ALAC facility. Further, their testimony was to the effect that it was not the electric utility companies' obligation to understand the internal processes at ALAC; their obligation was to operate their generation and distribution systems in a manner that assured the safety of the general public, which included their customers. [36] Additionally, the trial court cited Weaver v. Valley Electric Membership Corporation, 615 So.2d 1375 (La.App. 2 Cir.1993) for the proposition that a power company is required to recognize and make reparation for a risk that takes effect through another's foreseeable negligence. Significantly, Weaver was an electrocution case that arose after the adoption of comparative fault in Louisiana; the "foreseeable negligence" referred to by the court was the negligent conduct of the plaintiff. Id. at 1384-1385. Weaver is simply inapplicable here, and the trial court erred in imposing a duty on the electric utility companies to recognize and make reparation for a risk that took effect through ALAC's and Big Three's negligence. The trial court also cited Calhoun v. Federated Rural Electric Insurance Company, 571 So.2d 672, 681 (La.App. 3 Cir. 1990), writ denied, 577 So.2d 14 (1991), for the proposition that the defendant utility company was liable when it failed to de-energize its line after it knew or should have known of defects present in the plaintiffs' equipment. The instant case is distinguishable. In Calhoun, the defect on the plaintiffs' premises was electrical in nature, the use of an extension cord as a conductor from the breaker box. In the instant case, the record shows there were no electrical malfunctions in the ALAC equipment, merely the allegedly dislodged metallic particles. [37] Although the court in Northern Assurance Company, 580 So.2d 351, 358, reviewed that case pursuant to the manifest error standard, we pretermit discussion of whether review de novo is appropriate once the duty is defined erroneously because we ultimately conclude that the trial court was manifestly erroneous. [38] Ultimately, the electric utility companies' breach of the lesser duty of reasonable care is inconsequential in light of our later determination that plaintiffs' case fails under the cause-in-fact and legal cause inquiries. [39] This court may not disturb the conclusions reached by the trial court regarding factual matters in the absence of "manifest error" or unless the particular finding of fact was "clearly wrong." Rosell v. ESCO, 549 So.2d 840, 844 (La.1989). A two-part test is used in making this determination. This court may only reverse the trial court if: (1) we find from the record that a reasonable factual basis does not exist for the finding of the trial court, and (2) we further determine that the record establishes that the finding is clearly wrong or manifestly erroneous. Stobart v. State, Department of Transportation and Development, 617 So.2d 880, 882 (La.1993). Although deference to the fact finder should be accorded, this court has a constitutional duty to review facts. Of course, we may not merely decide if we would have found the facts of the case differently. Thus, we should affirm the trial court where its judgment is not clearly wrong or manifestly erroneous. Nevertheless, because this court has a constitutional function to perform, we have every right to determine whether the trial court finding was clearly wrong based on the evidence, or clearly without evidentiary support. Ambrose v. New Orleans Police Department Ambulance Service, 639 So.2d 216, 221 (La. 1994). Furthermore, when findings are based on determinations regarding the credibility of witnesses, the manifest error/clearly wrong standard demands great deference to the trier of fact's findings. However, when documents or objective evidence so contradict the witness's story, or the story itself is so internally inconsistent or implausible on its face that a reasonable fact finder would not credit the witness's story, the court of appeal may well find manifest error or clear wrongness even in a finding purportedly based upon a credibility determination. Rosell, 549 So.2d at 844-845. [40] Consider an ordinary deck of cards which consists of four suits: hearts, diamonds, spades, and clubs. Simply because it is a "statistical possibility" that a heart will be randomly drawn from the deck of cards, that is not proof that when a card is actually drawn it will be a heart. A "statistical possibility" addresses what can occur statistically as opposed to what actually occurs. Testimony regarding what is a statistical possibility does not translate into proof of what more probably than not occurred. The burden of proof for cause in fact is by a preponderance of the evidence—what happened more probably than not, as opposed to what is possible from a statistical standpoint. "Proof which establishes only possibility, speculation, or unsupported probability does not suffice to establish a claim." Todd v. State, Department of Social Services, Office of Community Services, 96-3090, p. 16 (La.9/9/97), 699 So.2d 35, 43. [41] Ultimately, the weight to be given expert testimony is dependent on the facts on which it is based, as well as the professional qualifications and experience of the expert. For an expert opinion to be valid and to merit much weight, the facts on which it is based must be substantiated by the record. Barry v. Plaquemine Towing Corporation, 96-0979, p. 5 (La. App. 1 Cir. 8/4/97), 698 So.2d 1017, 1020, writ denied, 97-2579 (1/9/98), 705 So.2d 1102. If the expert opinion is based on facts not supported by the record, the opinion may be rejected. Id. [42] The fact it is not an unusual occurrence for particulate matter to find its way to the automatic pressure control valve is established by a comprehensive review of the Schmidt report. The report states that upon inspection of the automatic pressure control valve and piping from the accident site, "[e]vidence that foreign material had previously existed within the piping system was discovered.... [T]he LO-DB seat ring cage [part of the automatic pressure control valve] exhibited old areas of abrasion and gouge damage." Furthermore, "[v]isual inspection and metallurgical analysis of the [automatic pressure] control valve's surviving components confirm the long term presence of foreign material within the piping system." This observation established these conditions existed prior to the subject fire and explosion. The record reveals the automatic pressure control valve and the piping had not been cleaned in years. No doubt the lack of maintenance led to this long term condition described in the Schmidt report. The importance of pipeline maintenance was stressed in the Lowrie article, previously quoted, and in the article entitled "Safety Considerations Regarding the Use of High Pressure Oxygen," also previously quoted, which was an attachment to the deposition of Dr. Chester Grelecki. [43] As indicated, Mr. Mabile testified without contradiction that no debris was discovered until a few months after the explosion and that the debris was located in an intake valve. [44] For an extensive discussion of legal cause/proximate cause/scope of duty, see Justice Cole's opinion on rehearing in Roberts, beginning at page 1052. The court said in pertinent part: Because substandard conduct does not render the actor liable for all consequences spiraling outward until the end of time, the concept of proximate cause, or one of its functional equivalents, such as scope of the duty in duty-risk analysis [or legal cause], is necessary to truncate liability at some point. Roberts, 605 So.2d at 1052. [45] But see n. 52, infra for a discussion of this issue. [46] Mr. Hracek had 27 years of experience, Mr. Perkins had 9 years of experience, and Mr. Bujol had 5 years of experience. [47] The record indicates Mr. Perkins and Mr. Bujol participated in at least 30 full plant shutdowns and at least 90 partial shutdowns between them. The record is silent in that respect as to Mr. Hracek, but obviously, with his experience of 27 years, his participation no doubt exceeded those numbers. [48] Mr. Bujol testified he had a concern about pressure at the site of the automatic pressure control valve, but not a concern about an explosion. [49] Because the legal cause issue is fact bound, other cases serve only as examples for methodology. Roberts, 605 So.2d at 1055. Other cases can be analogized only when the facts bear a striking resemblance to the case to be decided. Id. The trial court cited Istre v. Fidelity Fire & Casualty Ins. Co., 628 So.2d 1229 (La.App. 3 Cir.1993), writ denied, 94-0073 (3/18/94), 634 So.2d 852, in making its legal cause determination. However, we conclude Istre does not provide a striking resemblance to the case before us. In Istre, there was a power outage, but it was caused by the negligence of a backhoe operator who knocked out power to a traffic signal while working for a construction company. One cannot compare the legal duty owed by a backhoe operator to the driving public with the legal duty owed by an electric utility company to its customer. Furthermore, although the time span between the initial event and the accident in Istre was several hours, the electrical outage was still in existence, making the effects of the negligence continuous until the accident. In contrast, the accident in the instant case did not occur until three hours after the power disturbance ceased. [50] We refer to the lack of maintenance on the pipeline, the long standing problems with the automatic pressure control valve, the failure to install filters which would have prevented the accident, and requiring Mssrs. Hracek, Perkins, and Bujol to trouble-shoot the malfunctioning valve without the benefit of protective walls or barriers and without depressurization or isolation. As then Judge Lemmon stated in Hall v. Hartford Accident & Indemnity Co., 278 So.2d 795, 799 (La. App. 4 Cir.1973), writ denied, 281 So.2d 753 (La.1973): When the party charged with the responsibility of observing safety factors fails to do so, it is grossly unjust to place the blame for a resulting accident on the person who poured the last cup of water before the defective dam broke, unless that person also exercises a substantial amount of knowledgeable control over the dangerous situation. [51] The Schmidt report states: "Many factors affect particle impact type ignitions and thus lead to event randomness." [52] Because we determine that the trial judge was manifestly erroneous regarding his legal cause determination, we pretermit the issue of what is the proper standard of review on appeal for legal cause. The interrelated issue of whether legal cause is a fact issue or a legal issue has baffled scholars and courts for decades and begs for resolution in Louisiana. See Robertson, 58 La.L.Rev. at 19: "The Louisiana Supreme Court has vacillated on whether legal cause should be a question of law [subject to de novo review] or a question of fact [subject to manifest error review], and the lower court decisions sometimes reflect the uncertainty. However, prevailing practice [at the trial level] treats the issue [of legal cause] as one for the trier of fact. (Footnotes omitted.)" Robertson cites Kenney v. Cox, 95-0126 (La.3/30/95), 652 So.2d 992, 992 (Dennis J., concurring): "I feel that our jurisprudence has not clarified the distinction between the existence of a general duty of care (a legal question) and the `legal cause' or `duty/risk' question of the particular duty owed in a particular factual context (a mixed question of law and fact)[.]" Also cited is Freeman v. Julia Place Limited Partners, 95-0243 (La.App. 4 Cir. 10/26/95), 663 So.2d 515, 517, writ denied, 95-2808 (1/26/96), 666 So.2d 680, which documents the vacillation. If legal cause, which essentially serves the same purpose as scope of duty or proximate cause, is "a mixed question of law and fact," it should be resolved by the jury at the trial level and then be subject to manifest error review at the appellate level. As indicated, Fowler, 556 So.2d at 4, listed five elements for evaluating negligence: the duty element is to be decided by the judge, but the other four elements of breach, cause in fact, scope of liability (legal cause) and damages, are to be decided by the jury unless reasonable minds could not differ. See also Wilson, 576 So.2d at 493. This division of labor between the judge and jury is essentially what is proposed in W. PAGE KEETON, ET AL., PROSSER AND KEETON ON THE LAW OF TORTS § 45, 319-321. "[T]he hornbook says proximate cause is a question for the jury. As such, one would expect a court to review a jury decision on proximate cause much as it would review any jury determination, with great discretion accorded the factfinder." Galligan, 53 La.L.Rev. at 1513. Consider Reed v. Wal-Mart Stores, Inc., 97-1174, pp. 3-4 (La.3/4/98), 708 So.2d 362, 364, holding that whether a thing creates an unreasonable risk of harm is a mixed question of law and fact and, thus, subject to the manifest error standard of review. See also LSA-C.C.P. art. 1812(C)(1)(a), which provides the jury is to answer the legal cause question. See also Robertson, 57 La .L.Rev. at 1104 and 58 La.L.Rev. at 34, wherein the author makes the argument for allowing the trier of fact (judge or jury) to decide legal cause issues. However, if legal cause is a question of law, it should be decided by the judge, as opposed to the jury, at the trial level and be subject to de novo review at the appellate level. See Todd, 96-3090 at 6-7, 699 So.2d at 38-39, wherein the Louisiana Supreme Court states unequivocally that legal cause is purely a legal issue. Todd never mentions Fowler or Wilson. See also Galligan, 53 La. L.Rev. at 1522-1525, noting Professors Green and Malone, the fathers of duty-risk analysis, indicated it was the obligation of the judge to determine whether the duty extended to encompass a particular risk. [53] Any suggestion that the testimony of Tim Ponseti, a senior account representative with Entergy Services, Inc., and an Entergy investigative report established ease of association is erroneous. Mr. Ponseti's testimony was that there was "the potential the plant or someone else may connect" the voltage sag with the explosion; the Entergy investigative report noted a "Possible ** Backlash" from the ALAC accident. The Ponseti testimony and Entergy report are an acknowledgement that someone "potentially" or "possibly" would connect the voltage outage with the explosion and are insufficient to establish ease of association. When considering legal cause, ease of association is between the duty and the risk, not between two factual events. See Sibley, 475 So.2d at 319. [1] I note that although the majority opined that the trial court applied an incorrect duty of care in this case, and, thus, committed, in the majority's eyes, a legal error, the majority, nevertheless, uses the language of a "manifest error" review of the case. Here, I differ on two points with the majority. First, I am of a different opinion as to which standard of care the electric utility companies should be held. Second, while I do not believe that a legal error was committed in this case, I believe that once this court determines that legal error has occurred, this court should review the entire case de novo. See Ferrell v. Fireman's Fund Insurance Co., 94-1252, p. 7 (La.2/20/95), 650 So.2d 742, 747 (Where one or more trial court legal errors interdict the fact-finding process, the manifest error standard is no longer applicable, and, if the record is otherwise complete, the appellate court should make its own independent de novo review of the record and determine a preponderance of the evidence.). [1] Speculation in defendants' brief on rehearing as to why this case was initially heard by a five-judge panel is erroneous. Although issues were discussed by the initial three-judge panel, no decision was made regarding any issues prior to the determination that this matter should be initially heard by a five-judge panel. Given the myriad of issues, it was believed that inevitably at least one issue could not be resolved by the three-judge panel. The Perkins plaintiffs are partially correct when in brief they write: "... this court desired to have a panel of more than three judges decide this very significant case." [2] The purpose of allowing an application for rehearing is to afford to litigants, whose rights and interests are affected by the judgment, an opportunity to have an erroneous judgment corrected by the court deciding the case. The evaluation of whether to grant such a rehearing is solely within the discretion of the court. Eicher v. Louisiana State Police, Riverboat Gaming Enforcement Division, 97-0121, pp. 5-6 (La.App. 1 Cir. 2/20/98), 710 So.2d 799, 803, writ denied, 98-0780 (5/8/98), 719 So.2d 51. [3] Oral argument in support of an application for rehearing is not permitted. Uniform Rules, Courts of Appeal, Rule 2-18.1. However, after a rehearing is granted, the court may order oral argument. Uniform Rules—Courts of Appeal, Rule 2-18.5. [4] The request to file supplemental briefs was denied by a four-to-one vote of the panel. [5] Although the plaintiffs cite Duffy for the proposition that they are entitled to re-argue their case, we note that the supreme court was divided on the entitlement issue. However, the issue arose in a case where the initial argument was before a three-judge panel and La. Const. art. V, § 8(B) clearly applied. Thus, Duffy is factually distinguishable from the instant case.