Cities Service Co. v. Ocean Drilling & Exploration Co.

813 F.2d 679

• Court: Court of Appeals for the Fifth Circuit
• Decided: April 1, 1987
• Docket: No. 85-3583
• Status: Published

Opinion

JOHN R. BROWN, Circuit Judge:

This case arose from a blowout and fire which occurred on board the OCEAN KING, a drilling barge, which killed five people and caused several million dollars in property damage. The Fifth Circuit confronts it for the second time. Cities Service Co. v. Ocean Drilling and Exploration Co., 758 F.2d 1063 (5th Cir.1985). We there remanded the case for the District Court to enter adequate findings of fact and conclusions of law in conformity with F.R.Civ.P. 52. Both parties contend that the findings of fact are inconsistent and present this Court with the task of explaining their consistency.

The District Court found fault on behalf of all parties in varying proportionate degrees, but only granted recovery against Hydril Company (Hydril) for the personal injury/death claims which were settled before trial and brought as assignees by Cities Service Company (Cities) and Ocean Drilling and Exploration Company (ODE-CO). Oddly, the Court expressly declined to grant recovery to Cities and ODECO for the physical damages and losses they suffered because the proportionate fault attributed to Hydril was small. Cities and ODECO appeal that action while Hydril appeals the Court’s finding that Hydril failed to give adequate warnings on the use of the blowout preventer (BOP).

We hold that the District Court adequately made and reconciled any conflicting findings of fact on which to base its assessment of fault and causation against Hydril. We reverse the District Court’s failure to apply pure comparative negligence principles and remand in order for the District Court to award recovery to Cities and ODECO for their property damages, including those sought as assignees, consistent with the declared proportionate degree of fault against Hydril.

In the Beginning

ODECO owned and manned the jack-up drilling barge OCEAN KING which was engaged in drilling a well near Matagorda Island off the Gulf Coast of Texas. Cities and Getty Oil Company were the owners of the lease, with Cities designated as the operator. The OCEAN KING was equipped with two types of blowout preventers. Hydril Company manufactured the Hydril GL 5,000 annular blowout preventer ¹ that was installed on the OCEAN KING, which, as reflected by the literature Hydril disseminated to the industry, had a rated capacity of 5,000 pounds per square inch (psi). The OCEAN KING was also equipped with ram blowout preventers² which had a rated capacity of 10,000 psi. Government regulations require weekly testing of all BOPs at 70% of their rated capacity. The Hydril 5,000 GL annular BOP used by the OCEAN KING was thus tested at 3,500 psi each week.

The Blowout

The blowout onboard the OCEAN KING occurred around 6:00 to 6:15 a.m. on August 30, 1980. At that time, the OCEAN KING had drilled to a measured depth of 13,794 feet and had installed casing to a depth of 10,337 feet. The events leading up to the blowout commenced the night before, when the drill crew on duty began removing the drill pipe from the hole.

By measuring the amount of drilling fluid pumped into the hole to replace the pipe, the drill crew discovered that the volume of drilling fluid filling the hole was less than the volume of pipe being removed from the hole. This indicated an intrusion of some foreign substance into the well bore. The crew pulled ten more stands of pipe³ and determined that the well was no longer taking any drilling fluid. The drill pipe was then returned to the bottom of the hole in an effort to circulate the fluid and remove the gas.⁴ The drill crew could smell gas mixed with the drilling fluid and identified a decrease in the weight of the drilling fluid which also indicated that gas was mixed with the drilling fluid.

Approximately fifteen minutes later, mud blew out of the dome nipple on top of the blowout preventer. The crew then closed the Hydril annular blowout prevent-er and measured the pressure at 2,000 psi. The pressure continued to build and at 3,700 psi the Hydril blowout preventer, which had a rated capacity of 5,000 psi, but a tested capacity of 3,500 psi, failed to function by releasing its seal around the drill pipe causing drilling fluid to erupt 60 to 70 feet above the drill floor. As a result of the loss of drilling fluid, more gas was able to enter the well bore and the casing pressure quickly climbed to 4,800 psi. The ram preventers were closed and the well was resealed.

Although the rig crew continued its efforts to control the well, the situation worsened. When the pressure reached 7,200 psi the casing ruptured, the well blew out, and the rig caught fire. This blowout and fire resulted in the death of five persons, serious burns and injuries to several more, severe damage to property of Cities and ODECO, physical damage sustained by third parties, well-control costs, and lost profits.

Prior to trial, Cities and ODECO settled the issue of fault in causing the blowout between themselves; they then settled all the personal injury and wrongful death claims of the original individual plaintiffs and took assignments of the claimants’ rights against the remaining defendants. The case then proceeded to trial on the liability issues with Cities and ODECO seeking recovery from Hydril for their property damage and as assignees of the personal injury and wrongful death claims.

The case was tried before Judge Beer with an advisory jury. The advisory jury returned a verdict allocating causation and fault as follows:

ODECO 55%

Cities 38%

Hydril 7%

100%

Special interrogatories were submitted to the jury which were answered, at least on their face, in an inconsistent manner. The advisory jury, recognizing the apparent inconsistency, attached an advisory note to its answers explaining the basis for assessing causation and fault against Hydril. This note, in part, stated:

This jury has concluded that although the rubber packing unit per se was not defective, the Hydril Company should not be held free of all liability because of what we feel are inadequate warnings of safe, maximum pressure levels.

Although the trial Judge approved the jury’s allocation of causation and fault, initially and on remand, he only allowed Cities and ODECO to recover against Hydril for the personal injury/death claims brought as assignees. He refused to allow Cities and ODECO to recover 7% of their own property damage claims from Hydril. On ODECO’s Motion to Amend Judgment, the Judge indicated that his failure to follow pure comparative negligence principles was based on the disparity in proportionate fault between the parties.⁵

When we remanded the case for the District Judge to enter specific Findings of Fact and Conclusions of Law, he was faced with the task of reconciling the conflicting findings for a second time. The Court’s Findings of Fact and Conclusions of Law on remand precisely followed his original disparate treatment of the personal injury/death claims and property damage claims. The Judge continued to differentiate between the personal injury/death claims and property claims on the basis of perceived inequities of the pure comparative negligence principles of admiralty law. In his Conclusions of Law on remand, the Judge stated:

Odeco and Cities are not entitled to recover against Hydril for their own property damage claims. Having resolved their differences, Odeco and Cities come before this court as a single entity seeking redress against Hydril____ Recovery by those 93% negligent parties against Hydril (7% negligent) is unacceptable. To allow recovery to an entity that is ninety-three percent at fault against an entity that is seven percent at fault sets up a process that amounts to a travesty.⁶

Neither party has been able to read the District Court’s Findings of Fact and Conclusions of Law in a wholly consistent manner and thus, has presented this court, as a dispassioned triumvirate, with the task of interpreting the District Court’s ruling.⁷ In order to explain the District Court’s findings, it is first necessary to restate those findings causing the parties such great consternation.

The District Court initially, and again on remand, expressly adopted the jury’s findings and allocation of causation and fault, including the 7% to Hydril, stating that the jury’s “collective insight ... has come as close as any entity could come to fairly and impartially allocating percentages of responsibility and because I agree, based on my own consideration of the case, with their findings.”⁸

The Doubt Instigator: Using the Right Tools Does Not Guarantee a Work of Art

The jury’s answers and handwritten note which were adopted by the Judge produced an apparent inconsistent verdict. This brings to mind what has been regarded as the outstanding advantages of special verdicts and the attendant burdens their use imposes on trial judges.

The special verdict permitted by Rule 49(a) is a splendid device for clarification of jury verdicts and for focusing the jurors’ attention on the disputed facts without the possible confusion that may result from a lengthy charge concerning the different legal rules that would apply if the jury reaches one factual conclusion rather than another. ... However, like all fine tools, it must be skillfully employed and its successful use requires the careful attention of counsel for all parties as well as of the court to be certain that the questions are framed to avoid the possibility of inconsistent answers.

Guidry v. Kem Mfg. Co., 598 F.2d 402, 405-06 (5th Cir.1979), cert. denied, 445 U.S. 929, 100 S.Ct. 1318, 63 L.Ed.2d 763 (1980).

The fact is that one of the sometimes unexpected, but wholesome results of special interrogatories jury submissions is to emphasize the absolute necessity that there be first a clear understanding of the precise legal issues for jury resolution and then a translation of them into articulate questions which may be authoritatively answered by a simple categorical. In a general way this is to say that not only is it the jury’s imprecision which is hidden by the traditional general charge and verdict. Many juridical errors of omission and commission by court and counsel are likewise perpetually concealed.

R.B. Co. v. Aetna Ins. Co., 299 F.2d 753, 756-57 (5th Cir.1962). The error uncovered by this extremely intelligent and hardworking jury was the failure of the Judge and counsel to “anticipat[e] all the likely results against the possible existence of conflict or the type or types of judgments which could be rendered under [the interrogatories].” Brown, J., Federal Special Verdicts: The Doubt Eliminator, 44 F.R.D. 338, 351 (1968).

In order to understand what the jury did — and what the Judge expressly adopted — the actual construct of the verdict form and the jury’s answers are important and require protracted analysis.

In the instructions on defect and product liability theory, the Judge was careful to emphasize that the giving of inadequate warnings was a product defect.⁹ The Judge was equally careful to emphasize that any defect (including inadequate warnings) must be a proximate cause of the accident to impose liability under either product liability or negligence. However, in the interrogatories, the Judge was not as careful and did not look on them as would a conscientious group of lay persons.

The verdict form was divided in three main parts, A, B, and C. Part A inquired as to the claims by Cities and ODECO against Hydril for their property damages.¹⁰ In Part B, the inquiry concerned reimbursement for payments made in settlement of claims of the original wrongful death and personal injury plaintiffs.¹¹ Finally, of decisive consequence was Part C which asked the jury to assign percentages of fault for the damages in Parts A and B caused by each party.¹²

Part A specifically asked the jury whether the property damage was caused by defective design or manufacture of the Hydril packing unit, but did not specifically list the third type of defect, inadequate warnings. The jury recognized the deficiency in the interrogatory as submitted with the collaboration of court and counsel and assured reconciliation of their answers by their handwritten addendum. Their handwritten note, quoted above, explained that they found no “per se” defect, but did find a defect in the warnings. In effect, the handwritten note which was adopted by the District Court amplified the advisory jury’s answer to Part A. The jury proceeded to find proximate cause between the defective warnings they identified in their handwritten note and the damages to the original personal injury/death claims in Part B and assessed the same proportionate fault for the personal injury/death claims and property damage claims in Part C.

[[Image here]]

Although the trial Judge approved the jury’s finding of causation and allocation of fault in interrogatory Cl, he only allowed Cities and ODECO to recover against Hydril for the personal injury/death claims brought as assignees. The Judge declined to allow Cities and ODECO to recover 7% of their own property damage claims from Hydril.

Cities and ODECO moved for amended judgment after the court entered judgment on the advisory jury’s findings and filed detailed memoranda supporting their motions. The trial Judge attempted to interpret or reconcile the jury verdict as indicating a different standard should be applied between the original claims and those brought on assignment.¹³ The trial Court rejected the parties’ arguments concerning the jury’s resolution of the incomplete interrogatories and indicated that the basis for his interpretation or reconciliation, which failed to follow pure comparative negligence principles, was the high percentage of fault collectively allocated to Cities and ODECO.¹⁴

We do not have to rely solely on the powerful but unexpressed inferences of the jury’s responses which F.R.Civ.P. 49(a) affords, because the Judge, in response to our directions remanding this case, made express Findings of Fact and Conclusions of Law to indicate the factual basis for his judgment. As we pointed out during the initial appeal of this case, because “an advisory jury does no more than advise the judge ... any errors relating to rulings before the jury and instructions to the jury need not be considered____ The district court is not bound by the [advisory] jury’s findings and is free to adopt them in whole or in part or totally disregard them.” Cities Service Co. v. Ocean Drilling and Exploration Co., 758 F.2d 1063, 1071 (5th Cir.1985) (citations omitted). Although we do not need to parse the advisory jury’s verdict for legal sufficiency, the trial Judge, as author of the interrogatories, revived the import of the jury’s responses when he accepted the jury’s findings and incorporated them in his own Findings of Fact.¹⁵

On remand, the District Court maintained its initial posture and allocated 93% of the total liability to Cities and ODECO because the Hydril BOP “should not have been used at pressures in excess of [the] 3.500 psi to which it had been tested”¹⁶ and allocated the remaining 7% to Hydril “for its failure to warn ODECO and Cities that users should not rely on the Hydril annular blow-out preventer at pressures above 3.500 PSI.”¹⁷

Hydril, asserting the trial court’s finding that the Hydril 5,000 GL BOP should not have been used in excess of 3,500 psi to which it had been tested, argues that recovery is precluded under the sophisticated user defense. Hydril further contends that similar negligence by the user and producer for the same injury and conduct is inconsistent with general negligence and specifically with products liability law regarding the duty to warn.

The Pressures of Inadequate Warnings

Product liability law of this Circuit clearly establishes that a manufacturer has a duty to warn consumers of dangers associated with the use of its product and the “lack of adequate warnings renders a product defective and unreasonably dangerous even if there is no manufacturing or design defect in the product.” Pavlides v. Galveston Yacht Basin, Inc., 727 F.2d 330, 338, 1985 A.M.C. 202, (5th Cir.1984) (footnote omitted); Martinez v. Dixie Carriers, Inc., 529 F.2d 457, 465-66 (5th Cir.1976). “The adequacy of the warning must be evaluated in connection with the knowledge and expertise of the user of the product____ [T]he supplier may rely on the professional expertise of the user and tailor its warnings accordingly.” Koonce v. Quaker Safety Products & Manufacturing Co., 798 F.2d 700, 716 (5th Cir.1986).

Hydril correctly argues that a manufacturer has no duty to warn a user of dangers when the user is already aware of that danger. See, e.g., Dougherty v. Sante Fe Marine, 698 F.2d 232 (5th Cir.1983) (seaman unable to recover for failure to warn when he knew the dangers of pulling the release lever of a lifeboat sixty feet above the water); Bakunas v. Life Pack, Inc., 531 F.Supp. 89 (E.D.La.1982), aff'd, 701 F.2d 946 (5th Cir.1983) (stunt man’s recovery barred under failure to warn because he knew of dangers associated with using an airbag with a rated capacity of 200 feet for a free fall of 320 feet); Hebert v. Brazzel, 403 So.2d 1242 (La.1981) (an experienced oil field worker denied recovery where a pressurized valve stem broke when two and a half times the normal torque was applied).

In the case at hand, the District Court did not find that Cities and ODECO knew of the danger from which the injuries resulted. The sophisticated user defense provides that a manufacturer has no duty to warn a user of dangers of which the user is already aware. Koonce v. Quaker Safety Products & Manufacturing Co., 798 F.2d 700 (5th Cir.1986). However, the sophisticated user defense applies only when the user knew of the particular danger, in this case, the unreliability of the BOP between 3,500 psi and 5,000 psi. The District Court’s finding of negligent use of a product does not necessarily imply, as is urged by Hydril, a finding of knowledge or that the manufacturer’s warnings were adequate.

Expert testimony indicated the practice in the industry was to switch to a higher rated BOP once the pressure rose above the testing pressure of the BOP being used at that time. This does not necessarily indicate (i) knowledge in the industry that the Hydril annular BOP was not reliable between the 3,500 psi to which it was tested and the 5,000 psi to which it was rated by Hydril, (ii) knowledge by Cities and ODECO of this fact, or (iii) Cities and ODECO did not reasonably need to be warned of this fact by Hydril. It does indicate a margin of safety accepted by the industry and violated by Cities and ODE-CO. Their violation of the industry wide safety margin constituted an ample basis for the District Court’s allocation of fault against Cities and ODECO.

It was imprudent for the personnel aboard the OCEAN KING to subject the BOP to well bore pressures in excess of the pressures to which it was tested each week. However, this imprudence does not necessarily indicate knowledge on behalf of Cities and ODECO, or the rest of the industry, that a blowout preventer could not and should not be relied upon at pressures greater than that to which it was regularly tested, yet less than its rated capacity. But in the days of comparative fault, Cities’ and ODECO’s failure to observe a claimed margin of safety does not cure the manufacturer’s defective warnings concerning the use of their product.

Hydril places great emphasis on the fact that the District Court did not expressly state that the BOP “failed.” The District Court’s finding stated that “[ajpparently the rubber packing unit opened at approximately 3,700 PSI, releasing its seal around the drill pipe, which apparently allowed drilling fluid to erupt some 60 to 70 feet into the air over the drill floor.” Findings of Fact, U 8. Offsetting this, Hydril posits that the BOP could have released its seal as a result of reduced hydraulic pressure supplied to the BOP by the rig accumulator — a machine which was not manufactured by Hydril. The District Court, after hearing conflicting evidence of Hydril’s tests showing the BOP held at 3,700-4,000 psi under scientific conditions and statements by eyewitnesses and shore personnel that the BOP gave out at 3,700 psi, expressly adopted the jury’s findings which included precisely the extent to which the property damages were caused by ODECO, Cities, and Hydril.¹⁸ Reversal is not required for two reasons. First, the District Court found factually that the BOP opened at pressures less than that to which it was rated — the equivalent of holding that it failed to function as expected — and made no reference to any evidence that the rig accumulator malfunctioned. Second, it imposed liability on Hydril solely for Hydril's failure to warn that the BOP should not be used at pressures above those to which it was tested.

The District Court expressly credited the testimony of William Hise, Cities and ODE-CO’s expert in petroleum engineering, reconstructing the well control situation. Hydril’s own expert also testified that it was serious error to rely on the Hydril annular BOP to hold more pressure than to which it had been tested. Based on this testimony the Court held that one of the ultimate causes of the blowout was the crew’s failure to transfer to the ram preventers once the pressures reached the 3,500 psi test pressure.¹⁹ When one reads the Court’s finding of fault against Hydril for failing to warn against the use of the BOP in this manner in conjunction with the finding that this use was a cause of the blowout, it becomes clear that Hydril’s failure to warn was held to be a proximate cause of the damage resulting from the use of the BOP at pressures above the testing pressure.

The District Court clearly found that Cities and ODECO had a greater degree of culpability than Hydril and assessed proportionate fault accordingly. Hydril’s failure to warn that the BOP could not be relied upon beyond the point to which it was tested created the defect in their product for which 7% liability was assessed by the advisory jury and thereafter by the trial Judge. This determination by the District Court is amply supported by the evidence and therefore accepted under the standard of review set forth by the Supreme Court in McAllister v. United States, 348 U.S. 19, 20, 75 S.Ct. 6, 8, 99 L.Ed. 20, 24 (1954) (“in reviewing a judgment of the trial court, sitting without a jury in admiralty, the Court of Appeals may not set aside the judgment below unless it is clearly erroneous”); see also, Chaney v. City of Galveston, 368 F.2d 774, 776 (5th Cir.1966).

Two Wrongs Don’t Make One Right

Having disposed of the factual differences between the parties’ interpretations of the District Court’s findings, we arrive at the question of whether, under United States v. Reliable Transfer, 421 U.S. 397, 95 S.Ct. 1708, 44 L.Ed.2d 251 (1975), it is legally acceptable for the trier of fact to find that each of two parties is guilty of the same character of negligence.

We answered this question affirmatively in Nutt v. Loomis Hydraulic Testing Co., 552 F.2d 1126, 1128 (5th Cir.1977). In that case, two vessels, each being negligently operated, collided. The court determined that the proximate cause of the collision was the negligence of both parties. Judgment was entered according to their proportionate fault and upheld by this court.

Therefore, it is legally consistent for Hydril, as well as Cities and ODECO, to bear the costs of this tragedy as a result of their respective failure to warn against using the Hydril BOP beyond pressures to which it is tested and for using the Hydril BOP at pressures above that to which it is tested. The conduct was a result of both parties separate negligence which can constitute separate liability for the same injuries. See Allied Chemical Corp. v. Hess Tank Ship Co. of Delaware, 661 F.2d 1044 (5th Cir.1981).

The Appeal That Really Wasn’t

Cities and ODECO appeal the trial Judge’s failure to apply pure comparative negligence in accordance with this Circuit’s precedent. Hydril’s response is that Cities and ODECO misstated the only issue that is before the court on appeal — Hydril’s cross-appeal which attacks the 7% fault assessed against Hydril as improper. While we find it difficult to understand how an appellant misstates the issue it is appealing because it differs with appellee’s cross-appeal, we agree that the issue of applying pure comparative negligence in maritime cases is well established.

After the filing of briefs, reply briefs, and well into oral argument, it finally came to light that Hydril does not really challenge Cities’ and ODECO’s contention on comparative fault. Counsel for Hydril candidly and correctly stated at oral argument that if this Court were to uphold the finding of 7% fault on behalf of Hydril as not clearly erroneous or barred by the sophisticated user defense, then the Supreme Court’s holding in Reliable Transfer requires us to impose liability against Hydril for 7% of all damages claimed by Cities and ODECO, including the property damage claims.

Reliable Transfer clearly established pure comparative negligence as the rule in maritime cases. The District Court’s failure to impose liability in proportion to fault clearly violates the rule of Reliable Transfer and the Fifth Circuit’s application of this rule. See, e.g., Lewis v. Timco, Inc., 716 F.2d 1425, 1984 A.M.C. 191 (5th Cir.1983) (en banc), 736 F.2d 163, 1985 A.M.C. 1185 (5th Cir.1984); Loose v. Offshore Navigation, Inc., 670 F.2d 493, 502 (5th Cir.1982) (“active and passive negligence have no place in a liability system that considers the facts of each case and assesses and apportions damages among joint tortfeasers according to the degree of responsibility of each party”); Gator Marine Service Towing, Inc. v. J. Ray McDermott & Co., 651 F.2d 1096, 1984 A.M.C. 1927 (5th Cir.1981); Leger v. Drilling Well Control, Inc. 592 F.2d 1246 (5th Cir.1979).

An advisory jury as well as the District Judge have determined the facts that are not clearly erroneous. Our function is not to provide three more opinions on the facts, but merely to review those facts under the clearly erroneous standard of Anderson v. Bessemer City, and McAllister v. United States. We have concluded that the fact finder’s determinations are not clearly erroneous, but on those facts reverse and remand to the District Court to award to Cities and ODECO 7% of their property damages, the personal injury and wrongful death settlements brought on assignment, and the costs incurred.

AFFIRMED IN PART, REVERSED IN PART AND REMANDED.

1 . A blowout preventer is a casing head control which is "designed to prevent the uncontrolled flow of fluids from the well bore by closing around the drill pipe or completely sealing the hole in the absence of drill pipe.” 8 H. Williams & C. Meyers, Oil and Gas Law 78. An annular blowout preventer is designed to securely close off the space between the casing and the drill stem, called the annulus, by applying hydraulic pressure to the interior rubber packing unit — a collar-like gasket. The packing unit seals around the drill pipe and prevents the escape of gas and drilling fluid from the well bore. See generally, Sutton Drilling Co. v. Universal Insurance Co., 335 F.2d 820, 822 and n. 5 (5th Cir.1964) (discussing blowouts and blowout preventers).

2 . A ram blowout preventer functions by completely closing off the entire opening of the casing when the drill pipe is not in the hole or to sever the drill pipe and seal the entire casing. Williams & Meyers, supra n. 1, at 720. The ram BOPs on the OCEAN KING could withstand greater pressures than the annular BOP, but restricted the options available to the drilling crew to expurgate gas from the well bore.

3 . A stand of pipe consists of two joints of drilling pipe approximately forty to forty-five feet in length.

4 . “The circulation of drilling fluid is accomplished by pumping drilling fluid down through the drill pipe to the bottom of the well bore where it exits from the bottom hole assembly. The drilling fluid thereafter returns to the surface through the annulus or the space between the drill pipe and ... the casing or the surrounding geological formation. Eventually, the drilling fluid is returned to the mud tanks on the rig____ By circulating the drilling fluid in this fashion, the drilling crew attempts to circulate out of the well bore a bubble of gas which has intruded into the well bore from the surrounding geological formation.”

In re: Incident Aboard D/B OCEAN KING on August 30, 1980, MDL No. 481, slip op. at 3 n. 1 (E.D.La. August 23, 1985) (on remand).

5 . See note 14 infra.

6 . In re: Incident Aboard the D/B OCEAN KING on August 30, 1980, MDL No. 481, slip op. at 8 (E.D.La. August 23, 1985).

7 . Both parties naturally construe the Findings of Fact in a light most favorable to themselves and argue that any findings against them are inconsistent and should be set aside. Therefore both parties contend that a decision by this Court that is adverse to their position must be based on a finding that the District Court’s findings were clearly erroneous under the standard set forth in Anderson v. Bessemer City, 470 U.S. 564, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985) and McAllister v. United States, 348 U.S. 19, 75 S.Ct. 6, 99 L.Ed. 20 (1954). Before examining the District Court’s findings under the clearly erroneous standard, we must first determine what the District Court found.

8 . Findings of Fact (f 18, In re: Incident Aboard D/B OCEAN KING on August 30, 1980, MDL No. 481, slip op. at 7 (E.D.La. August 23, 1985).

9 . The Judge instructed the jury that inadequate warnings constituted a defect as follows:

In order to prevail in this claim then ... ODECO and Cities must prove by a preponderance of the evidence that the product was defective, that is, that that product was unreasonably dangerous for normal use and further that the defective condition of the product, if indeed they prove that there was a defective) condition, was a proximate cause of the damages.

The instructions and warnings given by a manufacturer with respect to the capability and limitations that its products are part of the overall design. A manufacturer of a product must give warning of any danger inherent in the product made by him or any use of which he knows or should know in which the user of the product would not ordinarily discover. The warning should be such that if told it would make the product safe for the user. The duty of the manufacturer to warn, however, goes only to those dangers which are not obvious. A manufacturer is not compelled to warn knowledgeable purchasers of the dangers of which they either knew or should be aware.

If you find that the product was not defective under the law as I have instructed you, you need go no further. The verdict will be for Hydril on the products liability claim.

If you, however, find the product was defective, you must then go on the next question which is whether or not that defect was a proximate cause of the accident.

This accords with established law which holds that a product is unreasonably dangerous and therefore defective when the lack of adequate warnings renders the product responsible for the damage. LeBouef v. Goodyear Tire & Rubber Co., 623 F.2d 985 (5th Cir.1980); Perkins v. Emerson Electric Co., 482 F.Supp. 1347 (W.D.La.1980); Chappuis v. Sears Roebuck & Co., 358 So.2d 926 (La.1978).

10 . PART A

There are two separate types of claims involved in this case with which you must deal. The first are claims by Odeco and Cities against Hydril for property damages. Odeco and Cities contend that these property damages were caused by the defective design or manufacture of the Hydril packing unit.

11 . PART B

Secondly, Odeco and Cities seek reimbursement or contribution from Hydril for monies each expended in executing settlements with the original plaintiffs____

1. Do you find that any of the parties, Odeco, Cities or Hydril, was at fault, either in negligence or in products liability, in a manner that was a proximate cause of the damages suffered by the original plaintiffs?

12 . PART C

If you have indicated in Parts A and B that one or more of the parties (Odeco, Cities, Hydril) was negligent, or manufactured a defective product, that was a proximate cause of the losses, I ask you to answer the following questions:

1. To what extent, expressed as a percentage, were the property damages described in Part A caused by:

Odeco 38%

Cities 55%

Hydril 7%

(emphasis added)

13 . Hydril, again pressing its contention of sophisticated users, attempted to reconcile this different standard on a distinction between the sophistication of the companies as users and the workers as users. The Judge dismissed this reconciliation, calling Hydril’s response "ingenious ... though perhaps not as legally meritorious" and reasserted the distinction based on the equities of allowing recovery when proportionate fault is widely imbalanced. Hearing on Motion to Amend Judgment, February 2, 1983 at 22.

14 . At the hearing on ODECO’s Motion to Amend Judgment, the Judge requested further briefing on the equities of applying pure comparative negligence where there is great disparity in the parties proportionate fault. Without ever couching them in terms of causation or lack of causation, the Judge stated his reasons for treating the original claims different from those on assignment as follows:

I should say for any reviewing court that there is little doubt in my mind that I would have done just that [applied pure comparative negligence principles] had the numbers been different. It is the numbers themselves that are having the overall'affect on me.

I have ascribed a different thesis for the ultimate determination of the property damage claims of the ninety-three percent negligent entities because of the numbers, because of my feeling that an equitable consideration dictates that entities that combine to be ninety-three percent negligent don’t have a basis for seeking recompense from the entity that was found to be seven percent negligent with respect to the claims of the plaintiffs.

Record, February 2, 1983 at 17-18, 29-30.

15 . The District Court's Findings of Fact on remand stated in part:

18. The advisory jury answered special interrogatories, supplemented by a handwritten note.3/ [Footnote 3 quoted the jury’s advisory/explanatory note] The interrogatories and accompanying note allocated responsibility for the blow-out and fire as follows: 38% to ODECO, 55% to Cities (and Getty) and 7% to Hydril. I accept these findings, being of the view that the collective insight of this jury has come as close as any entity could come to fairly and impartially allocating percentages of responsibility and because I agree, based on my own consideration of the case, with their findings.

19. Hydril was 7% at fault for its failure to warn ODECO and Cities that users should not rely on the Hydril annular blowout preventer at pressures above 3,500 PSI. (See Footnote 3).

Findings of Fact at 7.

16 . Paragraph 12 of the District Court’s Findings of Fact states:

Evidence was presented on the issue of whether it was proper to rely upon the 5,000 psi annular blowout preventer beyond the 3,500 psi pressure to which it had been tested. Ode-co and Cities witnesses ... as well as counsel for Odeco ... agreed that this issue was not really in dispute. The Hydril 5,000 psi annular preventer should not have been used at pressures in excess of 3,500 psi to which it had been tested.

In re: Incident Aboard D/B OCEAN KING on August 30, 1980, MDL No. 481, slip op. at 4-5 (E.D.La. August 23, 1985).

17 . Findings of Fact fl 19, Id.

18 . See notes 12 & 15 supra.

19 . The District Court summarized the testimony of Cities’ and ODECO’s petroleum engineering expert witness reconstructing the well control situation in paragraph 18 of its Findings of Fact following remand.

He testified that the ultimate cause of the casualty was the bursting of four successive casing strings in the well, the first of which being the 9Vs" casing that ODECO’s Mr. James examined after the casualty and described as extremely thin at the point of burst, that the rig crew should have vented the well before midnight, that the crew should have been evacuated by ODECO and Cities, that the casing used was of too light a grade, that the crew should have recognized the problem at the outset and initiated kick control procedures at 7:30 p.m. on the evening of the 29th, and that the crew should have transferred from the annular preventer to the rams before the pressure reached the 3,500 PSI test pressure. (Citations to transcript omitted.)