Justiss Oil Co. v. Kerr-McGee Refining Corp.

UNITED STATES COURT OF APPEALS

for the Fifth Circuit

_____________________________________

No. 94-41310

_____________________________________

JUSTISS OIL COMPANY, INC.,

Plaintiff/Cross-Defendant/

Appellant,

VERSUS

KERR-MCGEE REFINING CORPORATION,

Defendant/Counter-Claimant/

Appellee.

______________________________________________________

Appeals from the United States District Court

for the Western District of Louisiana

______________________________________________________

February 28, 1996

Before JOLLY, DUHÉ, and DENNIS, Circuit Judges,

DUHÉ, Circuit Judge:

Kerr-McGee Refining Corporation (“Kerr-McGee”) owns and

operates an oil refinery in Cotton Valley, Louisiana. In early

1990, the aluminum floating roof of a crude oil storage tank ("Tank

29") at the Cotton Valley refinery collapsed.1 Kerr-McGee

1

Tank 29, built in 1938, was a 55,000 barrel crude oil storage

tank. The tank was 114 feet in diameter and approximately 30 feet

high. It was accessible either by ascending a ladder leading to

the man way on the tank’s roof or through a side man way located a

few feet above ground level. In 1979, a cone-shaped aluminum

floating roof was installed inside the tank to act as a barrier to

the escape of vapors emitted by the materials stored therein. The

floating roof was damaged beyond repair, collapsed, sank into the

tank, and became partially submerged in the contents at the bottom

of the tank. The collapsed roof came to rest in the bottom of the

tank at an angle slanting upward from one side that actually

contacted the floor to the other side that remained about four (4)

contacted Baker Tank Company, a division of Justiss Oil Company,

Inc. ("Justiss"), and solicited a bid for the manufacture and

installation of a new floating roof. In response, Ken Moose, Baker

Tank Company’s Construction Manager, telephoned Gerald Collins, the

Kerr-McGee Plant Manager. During this conversation, Collins

requested that Moose submit a separate proposal for dismantling

and removing the collapsed roof from inside the tank. Moose agreed

to submit this bid, and asked to view the job site.

After meeting with Collins, the two men went to Tank 29 and

were joined by Herschel Jones, Kerr-McGee’s Maintenance Foreman.

Moose climbed to the top man way and looked inside the tank to

evaluate its condition and the floating roof. Kerr-McGee had

removed most of Tank 29's contents and was continuing to wash and

to vacuum the inside of the tank. Moose was satisfied that Baker

Tank Company could remove the roof, and upon Collins’ indication

that they had the job, agreed to deploy a crew to the refinery.

Over the three days that followed Moose’s visit, Kerr-McGee

continued to wash Tank 29 and utilized high-pressure steam to free

it of hydrocarbon vapors.

Subsequently, Justiss and Kerr-McGee contracted for the

removal of the floating roof.2 Under the contract, Justiss was

feet off the tank floor and above the side man way.

2

The district judge found the Kerr-McGee Construction or Field

Services Agreement was the contract between the parties. Def.’s

Ex. 17. Additionally, he found this agreement incorporated the

Baker Tank Quotation, submitted by Baker Tank as a bid for doing

this work, and the Terms, Conditions and Contractual Obligations

document accompanying the quotation. Pl.’s Exs. 21, 23.

2

obligated to provide all labor, supervision, materials, tools,

equipment, and services needed to complete the job. Further,

Justiss warranted that the work would be performed in a good, safe,

workmanlike manner in conformity with the highest industry

standards.3 Kerr-McGee was obligated to first "clean[] and gas

free[]" the tank, and to test the concentration of hydrocarbon

vapors in the tank's atmosphere each morning before the Justiss

crew entered.4 Although Kerr-McGee was only required to test the

tank once every morning, Justiss could request additional testing

at any time.5. “`[C]onfined or enclosed space’ means any space

having a limited means of egress, which is subject to the

accumulation of toxic or flammable contaminants or has an oxygen

deficient atmosphere. Confined or enclosed spaces include . . .

storage tanks . . . .” 29 C.F.R. § 1926.21(b)(6)(ii) (1989)

(submitted into evidence as Def.’s Ex. 39). 6. This term refers to

3

These obligations were imposed by the Kerr-McGee Construction or

Field Services Agreement, paragraphs 1 and 5, and Appendix A.

Def.’s Ex. 17.

4

The Baker Tank Quotation provided: “Tank to be cleaned and gas

freed by others [i.e., Kerr-McGee].” Pl.’s Ex. 21. Likewise, the

Terms, Conditions and Contractual Obligations document, in the

section titled “Tank Repair Work (Hot Work),” required that prior

to any on-site work by Baker Tank Company, Kerr-McGee must ensure

that the following work has been completed:

Tank is emptied, cleaned, decontaminated and freed of all

product, hazardous material, toxic and explosive gasses, and

is maintained at all times in a safe condition.

All pipe lines are disconnected and/or blanked.

Pl.’s Ex. 23.

5

Tank 29 was a “confined space.” See 29 C.F.R. §

1926.21(b)(6)(ii) (1989) (submitted into evidence as Def.’s Ex.

39). Industry standards require, for safe operations in a confined

space

like Tank 29, that the space be gas free.

The term “gas free” is defined in the industry as a lower

3

the use of a portable Combustible Gas and Oxygen Alarm to sample

atmosphere for combustible gases and vapors.7

The Justiss crew arrived at the refinery on Tuesday, March 6,

1990, and reported to the Kerr-McGee offices. While there, Darwin

Parker, Manager of Operations and Safety at the refinery, gave the

crew foreman, Thomas McKelvey, safety pamphlets that he and his

crew were to read and to sign prior to commencing work. After all

the signed pamphlets were collected,8 Parker took the Justiss crew

to Tank 29. Another Kerr-McGee employee sniffed the tank, obtained

an L.E.L. reading of less than 10%, and issued a hot work permit

authorizing the crew to go to work.

The Justiss crew worked for three and a half to four days

without incident. Each morning before the Justiss crew entered

Tank 29, a Kerr-McGee employee sniffed the tank, obtained an L.E.L.

reading of less than 10%, and gave McKelvey a hot work permit.

McKelvey signed each hot work permit verifying that all necessary

precautions had been taken and that work could begin.

On the fourth day of work, Sunday, March 11, 1990, Kerr-McGee

issued a hot work permit to McKelvey around seven o'clock in the

morning that was effective until five o'clock that evening,

provided conditions inside the tank did not change. McKelvey's

crew consisted of three laborers, Steve Vega, Chris Doyle, and

Ricky Martin. After working all morning, the crew broke for lunch,

returned about an hour later, reentered the tank, and commenced

working again. Shortly thereafter, a Kerr-McGee employee charged

with the responsibility of refueling the air compressor that

powered the ventilation fan bolted to the side man way of Tank 29

asked McKelvey if the crew would take their afternoon break early

so he could refuel the compressor and go home. McKelvey agreed and

the crew took a fifteen to twenty minute break. When they

returned, Vega, Doyle, and Martin went back into Tank 29 via the

top man way while McKelvey went to his truck some seventy to one

hundred feet away to retrieve a can of gasoline. Before McKelvey

explosive limit (“L.E.L.”) reading of 10% or less. See Cleaning

Petroleum Storage Tanks, API Publication 2015, §§ 2.2.1, 2.2.3 (Am.

Petroleum Inst., 3d ed., Sept. 1986) (submitted into evidence as

Pl.’s Ex. 221). To obtain such a reading, the tank must be

“sniffed.” “Sniffing” refers to the use of a portable Combustible

Gas and Oxygen Alarm to sample the atmosphere for combustible gases

and vapors. If a “sniffing”

of the confined space indicates that the area is gas free, a “hot

work” permit may be issued. A hot work permit indicates that the

concentration of combustible gases

in a confined space like Tank 29 is at a level suitable for the

safe use of spark-producing tools over a designated period of time.

8

McKelvey testified that, although he signed the booklet

certifying he had read it, he in actuality had not. 25 R. at 145.

4

returned, while Vega, Doyle, and Martin were inside, Tank 29

exploded. All three men inside the tank were killed.9

After settling the claims of the deceased workers’ families,

Justiss sued Kerr-McGee on theories of tort and contract to recover

its losses. Kerr-McGee counterclaimed asserting breach of contract

by Justiss, and seeking damages for the destruction of its

property. A nine day bench trial was held. At the conclusion of

the trial, District Judge Donald E. Walter found that the source of

fuel for the explosion and flash fire was vapors emitted from

gasoline brought into Tank 29 by the Justiss crew. Accordingly, he

held Kerr-McGee free from fault. Further, he ruled in favor of

Kerr-McGee on its counterclaim. Justiss appealed, asserting five

errors.10

9

Dr. George M. McCormick, II, who autopsied the bodies, testified

that all three men died almost instantly. 22 R. at 114-15. More

specifically, Dr McCormick stated the cause of death for Chris

Doyle and Steve Vega was “[i]nhalation of flame, smoke and/or

superheated gas” resulting in acute cardiorespiratory failure, with

the concussive force of the explosion being a contributing factor.

As for Ricky Martin, the cause of death was “acute carbon monoxide

poisoning” from inhaling smoke, flame and/or superheated gas, also

resulting in acute cardiorespiratory failure. Pl.’s Ex. 144

(summaries of Dr. McCormick’s autopsy findings).

Additionally, both experts classified the fire in Tank 29 as

a “deflagration,” or flash fire. This is consistent with the

testimony of other witnesses on the scene that fire did not linger

after the explosion. See, e.g., 22 R. at 56-57 (testimony of

Herschel Jones); 25 R. at 111 (testimony of McKelvey that he did

not recall ever seeing any flames).

10

The specific errors asserted were:

(1) The Trial Court's application of "Occam's Razor", as a

controlling legal precedent is an erroneous and incorrect view

of the law and an abuse of discretion.

(2) The Trial Court's finding of fact, as to the source of

fuel for the fire, is not supported by substantial evidence

and is clearly erroneous.

(3) The Trial Court's favorable ruling on the Kerr-McGee

Counterclaim was based on a standard of proof far less than

required by law; i.e. a preponderance of the evidence.

(4) The Trial Court failed to follow and apply the doctrine

of strict liability.

(5) The Trial Court failed to follow and apply the law of

contract.

5

I. Error One

Justiss argues the district court erred by applying Occam's

razor as controlling legal principle in deciding which evidence of

the events leading to the explosion and flash fire in Tank 29 to

credit. More precisely, Justiss argues that the district court, in

relying on this premise, applied an incorrect legal standard in

making its factual findings.11 Justiss contends that we should

discard the district court’s factual findings and conduct de novo

review of the trial record in accordance with applicable law to

ascertain the true cause of these deaths.

In his oral findings, Judge Walter stated:

The testimony has suggested numerous theories as to how the

tank exploded. Clearly Tank 29 was not gas free at the time

of the explosion. The explanations of why the tank was not

gas free range from the very simple to indeed the hyper-

complex. I have used Occam's razor which is as valid

juridically as it is scientifically. Basically Holcomb's

[sic] razor is that the simplest of competing theories should

be preferred over more complex or subtle ones.

28 R. at 71-72.12 Though Justiss interprets this statement to

indicate that Judge Walter's fact-finding was confined by Occam's

Original Brief of Justiss Oil Company, Inc., at vi.

11

Justiss complained, "With due respect to the Trial Court, there

is no jurisprudential authority for the application of `Occam's

Razor' as a controlling juridical principle." Original Brief of

Justiss Oil Company, Inc., at 15. Continuing, Justiss argued,

"More important, the Trial Judge has based his entire ruling on an

unprecedented juridical concept, i.e. `Occams Razor'." Id.

12

Occam's, or Ockham's, razor has been defined as "the philosophic

rule that entities should not be multiplied unnecessarily."

Webster's Third New International Dictionary (Unabridged) 1561

(Philip B. Grove, ed. in chief, 1981). While Judge Walter's

interpretation may not fit squarely with this dictionary's

definition, his understanding does find support in the

jurisprudence. See, e.g., Commissioner of Internal Revenue v.

Engle, 464 U.S. 206, 230, 104 S.Ct. 597, 611 (1984) (Blackmun, J.,

dissenting); Brown v. Vance, 637 F.2d 272, 281 (5th Cir. 1981);

Alabama-Tennessee Natural Gas Co. v. Federal Power Comm'n, 359 F.2d

318, 335 (5th Cir.), cert. denied, 385 U.S. 847, 87 S.Ct. 69

(1966); Swann v. Olivier, 22 Cal. App. 4th 1324, 1329 ( Cal. Ct.

App. 1994); Swierupski v. Korn, 419 N.Y.S.2d 87 (N.Y. App. Div.

1979); Stockbridge Sch. Dist. v. Department of Pub. Instruction

Dist. Boundary Appeal Bd., 531 N.W.2d 624 (Wis. Ct. App.), review

granted, 537 N.W.2d 570 (1995).

6

razor, this interpretation is misguided. Judge Walter's oral

reasons make clear that his findings were not merely an exercise of

this 14th Century maxim, but were based upon his evaluation and

weighing of all evidence presented. In finding that Justiss' crew

was the source of the fuel, he stated:

This finding is obviously dependent on which McElvey [sic]

story I believe. That told to the E.M.S. workers or that

expressed to OSHA representatives and indeed in this court.

The E.M.S. workers had no reason to lie. The story makes

sense. And frankly accepting the rescue workers['] [sic] or

E.M.S. workers['] [sic] version of what McElvey's [sic]

admission was explains all. . . .

. . . Under any of [Justiss' several scenarios suggesting

how Kerr-McGee's conduct caused the fire], hydrocarbons could

have reached dangerous levels but [Kerr-McGee's] version is

more persuasive.

28 R. at 73. While the reference to Occam's razor may be

unfortunate for its potential to create post-judgment controversy,

Judge Walter's reference to the maxim is not inappropriate, because

in fact he did not abdicate his duty as fact-finder. Judge Walter

evaluated the evidence before him, weighed it according to his

determinations of credibility and trustworthiness, and drew

reasonable inferences and conclusions supported by his findings.

Thus, insofar as Justiss attacks the district court's findings as

made by the application of an allegedly improper legal standard, we

reject this contention and affirm Judge Walter's actions.

II. Errors Two, Three, and Four

Justiss’ assertions two, three, and four are grounded in a

single argument that the district court erred in finding Justiss

employees introduced into Tank 29 the source of fuel for the

7

explosion and flash fire. For example, in disputing the district

court's ruling on Kerr-McGee's counterclaims, Justiss states:

In the "Memorandum Ruling" issued on October 26, 1994,

the Trial Court found that the actions of Justiss/Baker Tank

Company were "grossly negligent". In his oral ruling the

Trial Court described the crew as "reckless". Such rulings

can only be predicated on the finding that the Justiss crew

brought gasoline into the tank, virtually insuring their own

demise.

Such a finding and the Trial Court's favorable ruling on

the Kerr-McGee Counterclaim are totally and completely

unsupported by the evidence. A complete review of the record

will show that such a finding is "clearly erroneous" and not

the "truth and right of the case." . . .

. . .

. . . The Kerr-McGee Counterclaim succeeds (or fails) on

the Trial Court's finding that the Justiss crew introduced

gasoline into the tank.

Original Brief of Justiss Oil Company, Inc., at 23, 25 (citations

omitted). Likewise, Justiss' argument regarding the district

court's failure to apply strict liability to Kerr-McGee's conduct

is summarized as follows:

The District Court considered opposing Motions For

Summary Judgment, regarding the application of Louisiana's

doctrine of strict liability to the fire in Tank 29. Upon

determining that the source of fuel for the fire was flammable

liquids brought into the tank by Justiss employees, the Court

never reached the issue. Justiss respectfully suggests that,

had the District Judge correctly resolved the question of a

fuel source, that is, found that the evidence clearly

preponderated in favor of a finding that the fuel for the fire

was the residual hydrocarbons left in the tank after Kerr-

McGee's failure to clean it, a correct application of

Louisiana law would have required the District Court to find

Kerr-McGee strictly liable for the damages suffered by

Justiss.

Id. at 28 (emphasis added). Thus, the decisive issue on appeal is

whether the district court committed clear error in finding that

Justiss employees provided the fuel source for the explosion and

flash fire that resulted in these deaths.

8

"Findings of fact, whether based on oral or documentary

evidence, shall not be set aside unless clearly erroneous, and due

regard shall be given to the opportunity of the trial court to

judge of the credibility of the witnesses." Fed. R. Civ. P. 52(a).

A finding of fact is said to be "clearly erroneous" when,

notwithstanding there is evidence to support it, the reviewing

court upon examination of the entire evidence is left with the

definite and firm conviction that a mistake has been committed.

United States v. United States Gypsum Co., 333 U.S. 364, 395, 68

S.Ct. 525, 542 (1948). This standard precludes a reviewing court

from reversing a finding of the trier of fact simply because it is

convinced that it would have decided the case differently.

Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct.

1504, 1511 (1985). In fact, "[i]f the district court's account of

the evidence is plausible in light of the record viewed in its

entirety, the court of appeals may not reverse it even though

convinced that had it been sitting as the trier of fact, it would

have weighed the evidence differently. Where there are two

permissible views of the evidence, the fact-finder's choice between

them cannot be clearly erroneous." Id. at 573-74, 105 S.Ct. at

1511. With this deferential standard guiding our inquiry, we turn

to the record.

Before addressing the district court’s contested factual

findings, we enumerate those facts which the parties agree are

established by the record: Justiss deployed an inexperienced,

untrained crew to remove this roof, and this crew worked for three

9

to four days in the residual sludge in Tank 29 without incident.

Either the ungrounded electrical junction boxes or the electric

extension cords employed by the Justiss crew, which entered Tank 29

by the side man way through a gap beneath the ventilation fan and

which were submerged at various times in the water and sludge on

the tank’s bottom, produced sparks that were more probably than not

the ignition source for the fire. Finally, Justiss admits that,

prior to the explosion, its crew reversed the direction of the

ventilation fan connected to Tank 29, so that the fan was blowing

air into the tank instead of suctioning air out, and that this

provided the oxygen necessary to sustain an explosion and flash

fire in the tank.

With these facts established, we note that three elements are

required for an explosion or fire to occur: (1) an ignition source;

(2) a fuel source; and (3) an oxygen source. See Cleaning

Petroleum Storage Tanks, API Publication 2015, § 2.2.1 (Am.

Petroleum Inst., 3d ed., Sept. 1986)(Pl.’s Ex. 221). Justiss

admits its employees provided the first and third elements. Thus,

with respect to the cause of the explosion and flash fire in Tank

29, we are concerned only with discovering who is responsible for

providing the fuel source.

Justiss argues the district court erred in not finding that

the fuel source was residual hydrocarbons left in Tank 29 by Kerr-

McGee as a result of Kerr-McGee’s inadequate efforts to clean the

10

tank.13 Continuing, Justiss claims the district court ignored

evidence preponderating in favor of this conclusion. First,

Justiss contends Kerr-McGee admitted a lack of specialized skill or

expertise in tank cleaning. Additionally, Justiss argues its

expert, H.G. Nebeker, a chemical engineer, testified that the tank

was not cleaned according to industry standards, that the most

likely source of fuel was residual hydrocarbons trapped either in

the sludge in the bottom of the tank or in the roof or its

pontoons,14 and that his examination of the results of chemical

13

Justiss originally claimed that Kerr-McGee also improperly

monitored Tank 29. The trial record establishes, however, that at

no time during the job did the Justiss crew request additional

monitoring. See, e.g., 25 R. at 87-88. Further, the parties

agreed that continuous monitoring would not have prevented this

accident because of the nature of the fire. See Reply Brief of

Justiss Oil Company, Inc. at 10-11; Supplemental Brief of Kerr-

McGee Refining Corp. at 2.

14

In this court, Justiss admits there are only two possible

sources for the fuel: residual hydrocarbons resulting from Kerr-

McGee’s failure to properly clean and gas free Tank 29 or gasoline

brought into the tank by Justiss employees. At trial, however,

Justiss attempted to establish two other possible fuel sources.

First, Justiss claimed Tank 29 had not been properly or

completely isolated from the rest of the refinery prior to the

explosion. Accordingly, because the refinery experienced an

“upset” on the day of the explosion, and because Tank 29 was a

“slop” tank prior to becoming incapacitated by the collapsed roof,

Justiss suggested that when Kerr-McGee’s employees reacted to this

upset by removing product from the refining process and returning

it to slop tanks, some of the product or its vapors entered Tank

29. This theory, however, was abandoned by Justiss’ counsel in

brief and at oral argument, when he admitted the record clearly

showed Tank 29 was properly and completely sealed off from the

refinery.

During examination of its expert, H.G. Nebeker, Justiss tried

to develop a second theory. Mr. Nebeker hypothesized that vapors

from oil resting atop a waste-water pond had backed up through the

waste-water drainage system and were being emitted into the air

surrounding Tank 29 from the ground drain located adjacent to the

tank. Because this theory was not set forth in Mr. Nebeker’s

pretrial report of his findings and opinions, the district court

11

analyses performed on sludge samples taken from the tank after the

incident was consistent with an explosion and flash fire fueled by

residual hydrocarbons. Justiss states that Kerr-McGee’s own expert

agreed that residual hydrocarbons were the fuel source, and

consequently, because only two experts testified, the weight of

such testimony was in Justiss’ favor.

To further support its position, Justiss complains that the

district court relied solely on the equivocal testimony of a single

witness to support its finding. Justiss argues E.M.T. John Byrd

never testified that McKelvey said he took gasoline into Tank 29,

and strenuously points out that McKelvey testified vehemently that

neither he nor any member of his crew ever took gasoline into the

tank. Moreover, Justiss argues that the district court not only

made an impermissible inference from Byrd’s testimony that gasoline

was taken into the tank, but also extended its error by inferring

from this finding that a vessel other than McKelvey’s gas can was

used to transport the gasoline inside. No evidence of another

container was ever provided, other than Kerr-McGee’s expert’s

speculation on its existence, and McKelvey’s gas can was proven to

have been found outside of Tank 29 intact and unscathed after the

explosion.15

sustained Kerr-McGee’s objection that the theory could not be posed

for the first time at trial and that the testimony was therefore

inadmissible. Justiss does not contest this ruling, and its

admission that Tank 29 was totally isolated from such external

flows would now be sufficient to discredit this view.

15

Justiss also complained that several key Kerr-McGee documents

pertaining to refinery activities on the day of the explosion were

never provided to them. Accordingly, Justiss argues a presumption

12

The record makes clear that the district court, in finding

that gasoline brought into the tank by the Justiss employees was

the fuel source, simply made a credibility choice. Judge Walter

favored the testimony of E.M.T. Byrd and rejected McKelvey’s

testimony as untrustworthy. Not only is such a determination not

clearly erroneous, but the theory accepted by Judge Walter is

supported by the evidence.

that unproduced evidence contains facts unfavorable to the party

failing or refusing to produce it arises in its favor and supports

its position. Herbert v. Wal-Mart, 911 F.2d 1044, 1046 (5th Cir.

1990) (citing Wigmore on Evidence § 285, at 192 (Chadbourn ed.

1970)).

As the district court noted, “[t]he spoilation [sic] argument

advances [Justiss] nowhere.” 28 R. at 108. The documents sought

by Justiss and not produced by Kerr-McGee consisted of the control

room log for March 11, 1990, tank summary reports for March 10 and

11, 1990, certain refinery gauge reports, and maps of the refinery

and tank farm indicating the layout of underground pipelines. The

relevance of such documents relates to the theory that the events

surrounding the plant upset on the day of the explosion provided

the source of fuel. See 26 R. at 18-20, 22, 64 (testimony of

Justiss’ expert, H.G. Nebeker). Because Justiss conceded this

theory was implausible at oral argument, we see neither relevance

to the contents of these documents, nor reversible error in Kerr-

McGee’s failure to produce them.

Also, Justiss relied on the testimony of Marvin Avant, a

vacuum truck driver for Hollingsworth Construction, to assert that

Kerr-McGee had Tank 29 vacuumed shortly after the fire to dispose

of the remaining sludge and to prevent further evaluation of its

explosive tendency. Avant testified that he recalled vacuuming

sludge containing large amounts of oil from Tank 29 on March 13,

1990, just two days after the accident. When queried by Judge

Walter, however, he indicated that his efforts at Tank 29 occurred

more like a week after the explosion. 23 R. at 188. Additionally,

Kerr-McGee contradicted Avant’s testimony with that of George

Jenkins, a contract laborer at the refinery in 1990 with whom

Avant stated he frequently worked. Jenkins testified that he and

Avant worked together all day on March 13, 1990, at places in the

refinery other than Tank 29. This testimony was corroborated by

Jenkins’ time ticket for March 13, 1990, (Def.’s Ex. 36) and

Avant’s testimony that he recalled working with Jenkins on that

date (23 R. at 186-87).

13

E.M.T. John Byrd’s testimony, taken in context with the

circumstances surrounding the explosion, establishes a sufficient

evidentiary basis to support the district court’s inference that

gasoline was present in Tank 29. Under examination by Justiss'

counsel, Byrd testified as follows:

Q: You do specifically recall questioning [McKelvey] about

what he would want to be doing carrying gas in the tank,

correct?

A: Yes, Ma'am.

* * *

Q: Did you phrase your questions in what [sic] terms of what

the hell?

A: I asked him -- When he said -- He told me that -- My

answer -- He told me that I came out to get a gas can. I

said, sir, what in the hell would want to make you carry gas

in a tank for?

The Court: What was his response?

The Witness: His response was that's what we were

cleaning our saws and saw blades with, your honor.

* * *

The Court: He didn't -- He didn't tell you

whether they had previously brought gasoline in there or not;

is that right?

The Witness: Only thing he said in that regard,

your honor, is that's what we were using to clean the gum off

our saws.

The Court: As I understand your testimony it

went like this. He told you he was going to get gasoline and

you said what the hell would you bring gasoline in there for

and he said that's what we -- were, past tense? That's what

we were cleaning saws and saw blades?

The Witness: That's what we were cleaning our saws

and saw blades with.

The Court: To get rid of the gunk?

The Witness: To get rid of the gunk, that was his

exact words.

The Court: That's it?

The Witness: That's it.

26 R. at 187-89.

14

This testimony is supported by McKelvey’s own testimony that

the saws and equipment used to dismantle the roof were never

removed from the tank, even at the end of a workday; that the saws

being used were designed to cut wood not metal and were burning up

at a rate of at least one a day; that on the days prior to the

explosion, he had gone to a nearby hardware store to purchase

replacement saws; that because the day of the explosion was a

Sunday, he assumed this store was closed so replacement saws could

not be purchased; and that throughout the job he regularly siphoned

gasoline from his truck to fuel the generator that powered the

tools, and in fact was returning to his truck to siphon more

gasoline at the time of the explosion and flash fire.

Additionally, James Starkey, owner of Bayou Ambulance Service and

a volunteer fireman on the scene, testified that he had a saw

passed out of the tank through the side man way to him “with a lot

of gunk on it.” 26 R. at 163. Starkey’s statement as to the

“gunk” is corroborated by a photograph of one of the saws removed

from Tank 29 that displays the saw engulfed in some substance and

the testimony of a Kerr-McGee investigator that the saws he viewed

appeared “covered with a black, dirty substance.” See Def.’s Ex.

58-1026; Pl.’s Ex. 235 at 89. Finally, E.M.T. Rita Byrd, who

responded to the scene with her husband, John Byrd, corroborated

her husband’s testimony that McKelvey repeatedly told them what the

crew was doing in Tank 29 was “damned stupid.” 26 R. at 190, 195

(testimony of John Byrd); 26 R. at 199, 20 (testimony of Rita

Byrd). Armed with McKelvey’s testimony, and the testimony of

15

unbiased witnesses such as James Starkey and the Byrds, the

district court had more than a sufficient factual basis from which

to infer that gasoline had been brought into Tank 29 by the Justiss

crew on the day of the explosion and flash fire.

The credibility of this conclusion increases when one

considers that the record establishes that the Justiss crew worked

for at least three days without incident. On those previous days,

the crew stirred around in the sludge, cut and dismantled the

floating roof, operated their tools and extension cords so as to

generate sparks, and disengaged the ventilation fan two to three

times a day for up to thirty minutes at a time to remove pieces of

the roof from the tank. Why then did not an explosion or fire

occur earlier if residual hydrocarbons escaping from the sludge

were the fuel source? The logical conclusion is that something

changed. The agitation of the sludge was no different. The

evidence, however, suggests, for the first time, the impetus to

introduce gasoline into the tank environment on that Sunday to keep

the irreplaceable saws operational.16

16

The only other alteration to Tank 29's environment or the work

procedures employed by the Justiss crew was the reversal of the

direction of the ventilation fan, such that it was now blowing into

the tank instead of suctioning air out. Dr. Otha J. Jacobus, Kerr-

McGee’s expert witness, testified that having the fan blow into

Tank 29 from the side man way at the bottom of the tank was highly

dangerous. Because hydrocarbon vapors are heavier than air and

thus low-lying in the tank, blowing air into the tank only served

to accelerate the speed at which these vapors would mix with the

tank’s atmosphere and produce an extremely volatile condition. 27

R. at 246-47; 28 R. at 37. The parties stipulated in the Pretrial

Order that the Justiss crew was responsible for reversing the fan.

7 R. at 1257.

16

The evidence also refutes Justiss’ remaining contentions.

First, is Justiss’ assertion that the expert testimony concerning

the source of fuel weighs in its favor. Kerr-McGee’s expert, Dr.

Otha J. Jacobus,17 testified that a small quantity of gasoline,

i.e., less than a quart, was consistent with the type of fire that

occurred in Tank 29.18 27 R. at 230-31. Dr. Jacobus also testified

that Nebeker’s theory of hydrocarbons in the sludge was

scientifically unsupportable. 27 R. at 229. Finally, Nebeker

17

Dr. Jacobus holds a Ph.D. in organic chemistry, which he defined

as “the study of molecules comprised principally of carbon and

hydrogen.” 27 R. at 199. He taught college courses focusing on

organic chemistry for almost twenty (20) years, belongs to numerous

research organizations in this field, and has published and

lectured extensively in the area. Dr. Jacobus has qualified as an

expert in over twenty (20) states and in several federal courts,

including the Western District of Louisiana. He has conducted a

number of investigations into chemical fires and explosions, is

intimately familiar with the OSHA regulations on confined space

entry procedures, and is capable of testifying as to the proper

methods of cleaning oil storage tanks. He is a chemist with

expertise in “chemical synthesis, analysis, chemical fires and

explosions, natural gas explosions, natural gas odorants and

products liability.” Id. at 202.

H.G. Nebeker was tendered by Justiss as an expert in “refinery

operations, refinery operating protocols, industry standards

insofar as refinery operations and protocols are concerned as well

[as] the economic evaluation of refinery equipment.” 25 R. at 222.

On cross-examination concerning his expertise, Nebeker admitted:

His company is not in the accident investigation business. In his

nine years with the company he has consulted on only two or three

accident investigations. He has never consulted in a refinery

explosion. He is not an expert in flash points, nor the vapor

concentration necessary to fuel an explosion. He is not a chemist,

nor an expert in chemical analysis. He is not an expert in L.E.L.

readings, tank cleaning, tank monitoring, or tank ventilation. 25

R. at 225-32.

18

Dr. Jacobus did testify that he considered two possible sources

of fuel to be likely: gasoline or some other cleaning agent carried

in by the Justiss crew or vapors entrapped within some entity in

the tank. 27 R. at 230. When asked by Judge Walter which source

he deemed most likely, Dr. Jacobus responded they were “equally

likely.” 28 R. at 38.

17

acknowledged that the sludge samples he reviewed were inconclusive

as to whether the sludge fueled the fire because they were taken

after the fire, and that gasoline could have been a possible fuel

source. 26 R. at 83, 88. Thus, the expert testimony is at least

equally weighted.

Additionally, the evidence offers an explanation of the

failure to find a vessel in which the gasoline was transported into

the tank. Dr. Jacobus stated that if something like a plastic

bottle or cup had been used, it most probably would have been

saturated with the gasoline and fully consumed by the fire.

Finally, the evidence supports the finding that Kerr-McGee

properly cleaned Tank 29. Dr. Jacobus testified that the

procedures followed by Kerr-McGee satisfied all applicable industry

standards, that Kerr-McGee’s efforts were effective despite the

collapsed floating roof, and that the alternative procedures

suggested by Nebeker were not only unnecessary, but also infeasible

and potentially dangerous.

With this evidence of fuel source before him, Judge Walter

found:

[H]ere is what I think happened on March 11, 1990. The

Baker Tank crew entered a gas free tank, 8 percent L.E.L. on

that Sunday morning. They used Skill [sic] saws throughout

the day. These saws were designed for cutting wood. The

extension cords were not secured from possible submersion in

the liquid sludge. Throughout the week these saws had ceased

to function, either burned out or failed in some manner

necessitating replacement. Two saws being purchased on March

6, a replacement saw on March 89 [sic], another replacement

saw on March 10. The gunk and sludge in the B.S. and W.

apparently clogged the saws causing the problem. On Sunday

the saws probably became inoperable again. It was Sunday. As

Mr. McElvey [sic] said, they couldn't purchase new saws. I

believe they used gasoline in the tank to clean the saw

18

blades. There were gasoline vapors present in the tank. The

ventilator fan was turned off for 15 minutes approximately and

then turned on before the men re-entered the tank. Blowing

in, the fan merely mixed the vapors with the tank air.

Whether from the Skill [sic] saw sparks, if they had time to

get to use them or from an extension cord shortage as has been

suggested or a lighted cigarette, the vapors ignited killing

all three Baker Tank men crew members in the tank. This

finding is obviously dependent on which McElvey [sic] story I

believe. That told to the E.M.S. workers or that expressed to

OSHA representatives and indeed in this court.

The E.M.S. workers had no reason to lie. The story makes

sense. And frankly accepting the rescue workers['] [sic] or

E.M.S. workers['] [sic] version of what McElvey's [sic]

admission was explains all. It is true Kerr-McGee was

obligated by the terms of the contract to provide and maintain

a clean and gas free working environment. The hot work permit

warranted that a specific L.E.L. reading had been obtained.

Baker Tank could have reasonably relied on those hot work

permits to represent that hydrocarbons had not been released

into the tank by any foreseeable means such as seepage, back

flows, seepage from the pontoons, [or] slop to the tank. But

that duty did not include responsibility for hazardous,

reckless behavior by the [Baker] Tank crew which raised the

L.E.L. percentages. . . .

. . . Under any of [Justiss' several scenarios suggesting

how Kerr-McGee's conduct caused the fire], hydrocarbons could

have reached dangerous levels but [Kerr-McGee's] version is

more persuasive. There was no breach. But for the gasoline

in the tank, the tank would have been gas free. Baker Tank

sent a crew that was completely untrained, unfamiliar with the

work at hand, the dangers involved or proper precautions.

Once on the job, the Baker Tank crew acted in a manner that

virtually insured an accident. . . . Mr. McElvey [sic] knew

one thing and one thing only. He had in his hand a hot work

permit and that was all he felt he had to know. He should

have also known he couldn't bring gasoline in there.

. . . It was Baker Tank’s conduct that caused the

explosion and resultant loss of life.

28 R. at 71-75. Thus, Judge Walter's finding that gasoline

introduced into the tank by the Baker Tank crew provided the fuel

source resulted from his weighing the evidence and electing to

credit the proponents of this theory.

In a non-jury trial, credibility choices and the resolution of

conflicting testimony remain the province of the judge, subject

19

only to Rule 52(a)'s clearly erroneous standard. Gifford v.

National Gypsum Co., 753 F.2d 1345 (5th Cir. 1985). Thus, "when a

trial judge's finding is based on his decision to credit the

testimony of one of two or more witnesses, each of whom has told a

coherent and facially plausible story that is not contradicted by

extrinsic evidence, that finding, if not internally inconsistent,

can virtually never be clear error." Anderson, 470 U.S. at 575,

105 S.Ct. at 1512. See also Port Arthur Towing Co. v. John W.

Towing, Inc. (In re Complaint of Port Arthur Towing Co.), 42 F.3d

312 (5th Cir.), cert. denied sub nom. Jarreau v. Port Arthur Towing

Co., No. 94-2122, 1995 WL 388028 (Oct. 2, 1995). Additionally, our

jurisprudence has held that the burden upon an appellant attempting

to show clear error "is especially strong where the findings are

primarily based upon oral testimony and the [district] judge has

viewed the demeanor and judged the credibility of the witnesses."

Bryan v. Kershaw, 366 F.2d 497, 499 (5th Cir. 1966), cert. denied

sub nom. Bryan v. Kershaw Mfg. Co., 386 U.S. 959, 87 S.Ct. 1030

(1967). See also 9A Charles A. Wright & Arthur R. Miller, Federal

Practice and Procedure § 2586 & n.5 (1995). Accordingly, the

district court’s factual findings “come here well armed with the

buckler and shield” of the clearly erroneous standard embodied in

Rule 52(a). Machinery Rental, Inc. v. Herpel (In re Multiponics,

Inc.), 622 F.2d 709, 723 (5th Cir. 1980) (quoting Horton v. United

States Steel Corp., 286 F.2d 710, 713 (5th Cir. 1961)). Thus, a

review of the entire record in this case does not leave us with a

"definite and firm conviction that a mistake has been committed,"

20

and so we cannot say Judge Walter's finding as to the source of

fuel was clearly erroneous.

III. Error Five

As to the fifth error alleged, Justiss failed to develop its

argument that the law of contract was not followed and applied by

the district court. This error is mentioned only in the Statement

of Issues section of Justiss' brief. When an appellant fails to

advance arguments in the body of its brief in support of an issue

it has raised on appeal, we consider such issues abandoned. See

Gann v. Fruehauf Corp., 52 F.3d 1320, 1328 (5th Cir. 1995); Green

v. State Bar of Texas, 27 F.3d 1083, 1089 (5th Cir. 1994).

Accordingly, we decline to address the merits of this issue.

IV. Conclusion

For the foregoing reasons, the decision of the district court

is AFFIRMED.

DENNIS, Circuit Judge, dissenting.

.

I respectfully dissent.

The trial court committed an error of law by its evident

assumption or determination that Kerr McGee owed no legal duty to

Justiss or its workers to protect them from the obvious

unreasonable risks of harm that were involved in the conditions

under which Justiss was proceeding to install a new floating roof

in Kerr McGee’s crude oil storage tank. The law is well settled in

21

Louisiana and elsewhere that a person is required to realize that

there will be a certain amount of negligence in the world, and when

the risk becomes serious, either because the threatened harm is

great, or because there is an especial likelihood that it will

occur, reasonable care may demand precautions against an

unreasonable risk of harm to another through the negligent or

reckless conduct of the other or a third person. Levi v. S.W. La.

Elec. Membership Co-Op (SLEMCO), 542 So.2d 1081, 1084 (La. 1989);

Weaver v. Valley Elec. Membership Co-Op, 615 So.2d 1375, 1383 (La.

Ct.App. 1993); Davis v. La. Power & Light Co., 612 So.2d 235, 236

(La.Ct.App. 1992), writ denied, 615 So.2d 336 (1993); Graves v. Lou

Ana Foods, Inc., 604 So.2d 150, 159 (La.Ct. App. 1992); Putt v.

Daussat, 381 So.2d 955 (La.Ct.App. 1980); Dragotis v. Kennedy, 250

N.W. 804 (Minn. 1933); Murphy v. Great Northern R. Co, 2 Ir. Rep.

301 (1897).

This basic principle of tort law is elaborated upon in the

authorities cited with approval in Levi v. SLEMCO, supra. For

example, Prosser and Keeton on Torts explains:

In general, where the risk is relatively slight, a person

is free to proceed upon the assumption that other people

will exercise proper care . . . . But when the risk

becomes a serious one, either because the threatened harm

is great, or because there is an especial likelihood that

it will occur, reasonable care may demand precautions

against “that occasional negligence which is one of the

ordinary incidents of human life and therefore to be

anticipated.” “It is not due care to depend upon the

exercise of care by another when such reliance is

accompanied by obvious danger.” . . . .

The duty to take precautions against the negligence

of others thus involves merely the usual process of

22

multiplying the probabiliity that such negligence will

occur by the magnitude of the harm likely to result if it

does, and weighing the result against the burden upon the

defendant of exercising such care. The duty arises, in

other words, only where a reasonable person would

recognize the existence of an unreasonable risk of harm

to others through the intervention of such negligence.

It becomes most obvious when the actor has reason to know

that he is dealing with persons whose characteristics

make it especially likely that they will do unreasonable

things.

W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 33,

at 198-99 (5th ed. 1984)(footnotes and citations omitted).

The Second Restatement of Torts provides that “[a]n act or an

omission may be negligent if the actor realizes or should realize

that it involves an unreasonable risk of harm to another through

the negligent or reckless conduct of the other or a third person.”

Restatement (Second) of Torts § 302A (1984). The comments under

this section are similar to those of Prosser and Keeton.

Significantly, comment c. states:

As stated in § 290, the actor is required to know the

common qualities and habits of human beings, in so far as

they are a matter of common knowledge in the community.

The actor may have special knowledge of the qualities and

habits of a particular individual, over and above the

minimum he is required to know, or he may have special

warning that the individual is or is about to be

negligent or reckless in the particular case.

Id., cmt. c. (quoted in pertinent part).

23

In the present case, Kerr McGee, particularly with its

superior knowledge, skill and experience in the storage and

refining of crude oil, should have recognized that the

circumstances and conditions under which the Justiss crew was

proceeding to install a new floating roof in the refinery company’s

crude oil storage tank involved an unreasonable risk of a

conflagration in a refinery endangering its own workers, the

Justiss employees and the public. The trial court described those

conditions and circumstances:

They used skill saws throughout the day. These saws were

designed for cutting wood. The extension cords were not

secured from possible submersion in the liquid sludge.

Throughout the week these saws had ceased to function,

either burned out or failed in some manner necessitating

replacement . . . . The gunk and sludge in the B.S. and

W. [a]pparently clogged the saws causing this problem .

. . . Whether from the skill saw sparks, if they had

time to get to use them or from an extension cord

shortage as has been suggested or a lighted cigarette,

the vapors ignited killing all three Baker Tank [a

division of Justiss] crew members in the tank . . . .

Baker Tank sent a crew that was completely untrained,

unfamiliar with the work at hand, the dangers involved or

proper precautions. Once on the job, the Baker Tank crew

acted in a manner that virtually insured an accident.

Three of the members had little, if any, prior training

in confined space work. The crew used skill saws

designed for wood use, they used gasoline to clean the

saws and blades. They used unsecured electrical

extension cords . . . .

I must say no one covered themselves in glory as far as

following simple -- Heck, following their own safety

regulations.

ER 6 (Oral Ruling of July 11, 1994).

The Justiss employees worked with the obviously hazardous

jerry-built electrical system in Kerr McGee’s crude oil storage

24

tank containing hydrocarbon residue located within the Kerr McGee

refinery for several days prior to the accident. During this time

the tank work site was inspected regularly and was continuously

subject to inspection by Kerr McGee employees. Consequently, Kerr

McGee knew or should have known that there was a serious risk of a

detonation or a conflagration caused by sparks from the skill saws,

the wiring or the extension connection boxes. The Baker crew left

its equipment in the tank overnight; the Kerr McGee inspection crew

could not have missed the electrical set-up or the type of

equipment that was being used in a tank considered dangerous enough

to sniff for gas fumes daily. Because the threatened harm was

great and there was an especial likelihood it would occur--as the

trial court put it, the circumstances “virtually insured an

accident” (ER 6)--reasonable care demanded that Kerr McGee take

additional precautions against the kind of accident that happened

by requiring safer work procedures, more frequent inspections of

the work site and monitoring of the Justiss crew’s work habits and

equipment. In fact, Kerr McGee owed this duty not only to the

Justiss employees but also to its own workers and the public as

part of its duty of reasonable care as a prudent oil refinery

operator.

The trial court evidently concluded that Kerr McGee owed no

duty to take precautions against any risk that might take effect

through a victim’s negligence. Instead, the trial court was of the

view that Kerr McGee was only under a duty imposed by contract to

sniff or test the air of the tank each morning to insure that it

25

was sufficiently free of gases to permit electrical work at that

time. The trial court thus erred as matter of law in not

recognizing and applying the law of Louisisana which imposes a

delictual duty to realize there will be a certain amount of

negligence in the world and to take precautions against an

unreasonable risk through the negligent or reckless conduct of a

victim or third person when the threatened harm is great or there

is an especial likelihood that it will occur. Consequently, the

trial court further fell into legal error in not considering to

what extent Kerr McGee may have been at fault delictually and in

not proceeding to quantify the degree or percentage of comparative

negligence attributable to each party. La. Civ. Code Ann. art 2323

(1995).+* The judgments of the trial court should be vacated and

the case should be remanded to it for the application of the

correct principles of law.

26