JOHN R. BROWN, Circuit Judge.

The question in this case is when is a blowout not a blowout. More accurately, the question is whether the occurrence damaging, if not completely destroying the well being drilled, amounts to a "blowout" as defined in the policy. Upon the jury verdict, the District Court determined that it was not a blowout and entered judgment for the Insurer. The Assured appeals. We reverse.

At the heart of the whole controversy is the policy definition of a blowout. Since that inevitably leads the parties and us into a comparison of this case with our decisions in Georgia Home Ins. Co. v. Means, 5 Cir., 1951, 186 F.2d 783, and Fidelity-Phoenix Fire Ins. Co. v. Dyer, 5 Cir., 1955, 220 F.2d 697, it is helpful to parallel the policy definitions in Means and Dyer. Unlike those cases, this one was tried to a jury whose verdict in response to special interrogatories and a general charge, F.R.Civ.P. 49(a) reveals precisely the determination of the subsidiary elements of the blowout definition. To the extent that any one or more of these represent questions of fact, it is necessary — if the Assured is to carry the day — that it convince us that its motion for instructed verdict and for j.n.o.v. based thereon, F.R.Civ.P. 50(b), should have been granted.

We defer for the moment any discussion of whether the facts gave rise to diverse inferences warranting jury submission to substantiate the verdict (note 4, supra). The trial itself was a short one in which for all practical purposes, the facts recited were uncontradicted. For our purposes, they may be briefly summarized and for this purpose we draw heavily on the briefs for this paraphrase.

During the month of April 1961, the Assured was engaged in the drilling of an oil well known as Sol West, Jr. No. 1, in Live Oak County, Texas. The well was to have been drilled to a total depth of approximately 10,000 feet.

On April 22, while drilling in a sand formation at a depth of 9,320 feet one of the crew members noticed that mud was increasing in the pits, indicating that more mud was coming out of the hole than was being pumped into the hole. As it was probable that the drill stem had encountered a subsurface formation which was exerting more pressure on the column of drilling fluid than the hydrostatic pressure was exerting on the formation, the tool pusher in charge of the drilling operation had the driller pull the drill stem, and the crew began to weight up mud in an attempt to control the subsurface formation pressure. The rate of return of the mud rapidly increased, however, and within about ten minutes the fluid began to flow up over the bell nipple in the top of the casing erupting approximately 10 to 15 feet above the derrick floor. As he feared a violent eruption of mud and gas, the tool pusher ordered the blowout preventer closed.

After closing the BOP, the crew continued to pump heavy mud into the hole, but rotation of the drill pipe could not be resumed because any attempt to turn it with the BOP closed would damage the rubber bushings and make it ineffective.

Efforts to equalize the circulation of the drilling fluid continued for a period of from four to six hours after the initial eruption of the drilling fluid from the well. Considerable heavy mud was pumped into the well, but it did not overcome the pressure imbalance from the subsurface formation. It was then observed that salt water was starting to come out of the choke line. The tool pusher concluded from this that the excessive pressure was forcing the heavy mud out into some formation down the hole somewhere, and that the well was cleaning itself of the drilling fluid and gas causing the walls of the hole to cave in.

All efforts on the part of the crew to restore circulation failed. They were apprehensive that the hole or bore of the well had bridged over causing the drill pipe to become stuck. Of course, no test could be made with the BOP closed. To determine the situation as to the drill pipe, the BOP was opened. Immediately a volume of mud in a column the diameter of the surface casing (10¾") was expelled out of the well approximately 20 feet into the air. With this expulsion continuing, unsuccessful attempts were made to rotate the drill pipe or move it. This proved what the tool pusher apprehended, that the drill pipe had become stuck because of the bridging over of the hole. The BOP was immediately reclosed. About this time, those in charge concluded that it was not possible to resume the previous drilling operation and the decision was then made to attempt to kill the well. Approximately 700 barrels of drilling mud costing $15,000 to $20,000 was pumped into the hole. This was ineffectual in overcoming the reverse pressure from the subsurface formation. It also indicated quite positively, from the loss of the fluid, that the mud-fluid was being forced out into one or more subsurface formations with undoubted damage to the wall of the hole. Efforts were then made to cement off the bottom of the hole. As we read the testimony, the initial hope was that the cementing would enable them to recover the drill pipe and then by accepted procedures, drill through or past and thereby complete the well substantially as originally planned. On the first attempt cement was pumped down the drill pipe, out through the bit, and up into the annulus. Because of extremely high pressures, which the Assured estimated to have been two to three times that normally encountered (4500 psi), the cement would not set at or near the bottom of the hole (approximately 9200 feet). The record indicates that at least one and perhaps two more attempts to cement were made. The last attempt successfully plugged the annulus at approximately 7300 feet. This kept the gas pressure from escaping in the annulus, but in the meantime the gas pressure was still being exerted inside the drill pipe. The hydrostatic pressure of the mud being pumped into the drill pipe and perhaps a surface valve prevented the escape of this gas (and fluids), but this condition could not be continued indefinitely. So it was decided to freeze the mud in the drill pipe by packing dry ice around the outside of it. With that frozen "plug" in the drill pipe, they were able to cut off the pipe, install the "Christmas tree," and complete it as a shut-in gas well. As a result of this incident, the Assured lost drill pipe, drill collars, and other undisclosed equipment of an agreed value of $33,266.62.

It is undisputed that the BOP did hold. When it was first closed, it functioned properly. When it was again closed after the attempt to turn or move the drill pipe, it held again. At all times, it prevented the escape of this gas (with liquids) from the annulus. That was its function. Mechanically it operated efficiently and successfully. It is likewise uncontradicted that although the Assured did not intend to bring in a gas well, or a gas well in this way, the well was finally brought under control by making a shut-in gas producer of it.

We must say that in our struggle with this case, we derive very little help from either Means or Dyer. The policy definitions are markedly different and each dependent on its own facts found by a Judge without a jury giving us a little more leeway under F.R.Civ.P. 52(a) than does the Seventh Amendment. But this factor is likely of little consequence. For we are equally convinced that this case does not turn on the sufficiency of the evidence to support the jury verdict. It turns on a matter of law as to the construction of this particular policy definition. This is highlighted by the fact that, as to some portions, the verdict cannot stand.

Thus, the answer to question No. 1, see note 4, supra, is contrary to the undisputed evidence. The jury in effect found that the expulsion of the fluid and gas was not "sudden." But under the trial Court's definition to which there was not objection, the term meant "happening without previous notice or with very brief notice; unforeseen; rapid. It does not mean instantaneously." On the evidence here, each and every one of these conditions is satisfied. It is true, of course, that the roughneck observed conditions indicating that something was going amiss. Within a few minutes, the tool pusher was summoned and the steps to control the gas were commenced. There is not the slightest suggestion that at any time prior to this initial discovery the Assured knew of, or should have foreseen, its likelihood in time to have taken other or different or more effective action.

The same thing is true by the jury's double negative answer to question (2), that expulsion of the fluid-gas was not uncontrollable. From a sheer physical point of view, it was "controllable" in the sense that the gas pressures were ultimately harnessed. But that result would be accomplished by plugging or capping the well. And yet no one would contend — certainly they have not here — that this would be "controllable." Again, a reasonable interpretation of this term is reflected by the trial Judge's instruction which met with the approval of all. Without foreclosing specific attacks on it when and as made below or in future cases, we think it is so inherently reasonable that at this stage it reflects a proper standard for evaluating the evidence. The Judge, as do we, used a practical approach by defining "controllable" in terms of steps capable of being taken to subdue the gas, but which would still permit the operators to make the well as planned. Closely related to this is question No. 3 as to which the jury found that the expulsion did not result in the well getting completely out of control.

As to each of these, the "expulsion" of fluid-gas and "the well" could be controlled by closing the BOP and keeping continuous hydrostatic pressure on the drill pipe. But so long as the BOP was closed, it was impossible to "continue to make the well" as the Judge's definition stated (see note 12, supra). Indeed, the momentary opening of the BOP to ascertain whether the drill pipe had been stuck resulted immediately in a further expulsion of gas-fluid in great volume and under high pressure. Likewise the inability to ever cement off both drill pipe and annulus made it impossible to "continue to make the well" or "to control the well" (see note 13, supra). Of course this approach assumes that "the well" relates to the well being drilled at the time the devastating natural forces began to operate. So defined, "the well" here was the projected well to 10,000 feet, not the one that was accidentally brought in as a shut-in gas producer at an unknown horizon below 7300 feet in the desperate, but successful, efforts to stop the flow of gas-fluid.

But questions (4) and (5) cannot be disposed of in this fashion. On the contrary, the overwhelming undisputed evidence would compel the factual answers given by the jury. The BOP properly performed its function when the sudden expulsion first commenced. It did so later after the BOP was momentarily opened in the attempt to turn the drill pipe. It continued to function until the annulus was cemented off after perforating the drill pipe. In a physical-mechanical sense, the BOP was therefore completely "operative" and completely "effective." That being so, neither the "expulsion" nor the "loss of control" rendered, i.e., made, the BOP inoperative or ineffective.

On these findings, the Insurer makes the very plausible argument that the language of the policy is plain and the facts are found. Therefore, it concludes, there is no "blowout" no matter how sudden, or uncontrollable the eruptive forces might be.

It then urges that the contract must be given literal effect and any other interpretation would offend the Texas law — by which we are Erie-bound — which forbids that a Court rewrite the policy.

However, we do not believe that it reasonably was intended that these words be so mechanically applied. Actually, we are not dealing with the "words" of element [3] of the definition clause, see note 1, supra. What we are dealing with is the word seldom encountered in insurance policies, "rendering" and its relation to the word "ineffective." On the Insurer's approach, the word "rendering" is read in terms of causation. On this reading for there to be a blowout there must be a conjunction of three things: first, a sudden uncontrollable expulsion which, second, results in the well getting completely out of control, and both of which, third, cause "the use of any blowout preventer equipment" to become "inoperative or ineffective."

That reading, of course, puts the whole emphasis on the mechanical result with respect to the BOP. It does not take into account any other kind of damage, no matter how devastating. It excludes damage even though occasioned by an uncontrollable expulsion of gas-fluid resulting in the well getting completely out of control if this happens without causing damage to or interference with the BOP. Indeed, as we understand the Insurer's argument, if the BOP worked perfectly, there would yet be no policy blowout even though the gas pressures restrained by the BOP caused bridging or caving making it absolutely impossible to ever recapture control of the well with the result that it would have to be abandoned and plugged. Likewise, there would be no recovery in the situation where, based on reliable geological information, there was almost certain production at 10,000 feet, but the pressures restrained by the BOP are so intense that cementing is successful only at a much higher level and production is obtained at, say, 3,500 feet.

We cannot believe that this is the purpose of the Insurer in the use of this unique transitional "rendering." We think that what is reasonably intended covers either one or all of these situations. First, where the conditions described in elements [1] and [2], note 1, supra, damages the BOP so that it is either inoperative or ineffective; or second, where, although the BOP is not damaged, the element [1] and [2] conditions are such that the BOP is unable to restrain the pressures; or third, with elements [1] and [2] conditions present, the circumstances are such that the BOP in working operating condition is not an effective means of avoiding the damage admittedly done by the extraordinary pressures.

Our case would be under the third reading. The damage was not due to operation or effectiveness of the BOP. On the contrary, the damage resulted despite the BOP. In other words, the operation of the BOP was utterly ineffective in the prevention of the damage done. The BOP was ineffective in regaining control of the well prior to the damage being sustained. It seems unlikely to us that where pressures are present (and do damage) which if unrestrained by an operative BOP would continue to satisfy all aspects of elements [1] and [2] the Insurer intends to exclude such damage from the blowout protection merely because in the perfect operation of the BOP, the BOP would not have prevented them. Those underwriting modern drilling risks surely do not seek irrelevant mechanical devices. The BOP is critical to the risk because, if working, it prevents damage. For those situations, the Insurer insists that the conditions either cause the BOP to fail, or make it ineffective. But where, although working perfectly, the damage from otherwise uncontrollable pressures can in no way be prevented by the BOP, we think the Insurer means to recognize that the BOP is ineffective.

On this interpretation of element [3] of the definition and our conclusion that with respects to elements [1] and [2], the verdict cannot stand, the usual result would be for us to reverse and either render or remand with directions to enter a judgment for the Assured. We think, however, that the unusual circumstances presented here call for a reversal and remand for a new trial as was done in Gulf Oil Corp. v. Wright, 5 Cir., 1956, 236 F.2d 46, 53.

This record reflects that the Insurer's defense rested essentially on element [3] and the literal application of it. That it was then justified in this reading gets considerable support from what appears to be a substantially parallel view of the trial Judge. Without a doubt, from the standpoint of both Assured and Insurer, our definitive reading of element [3] puts an entirely different color on the whole case. The Insurer should be free, of course, to try to meet our reading of element [3]. And both parties should be given an opportunity to try the case with respect to the other issues (elements [1] and [2]) with this basic policy interpretation of [3] in mind. It is not at all certain that more and different evidence, factual and expert, will not be forthcoming.

Of course this direction which we now give makes it more than ordinarily vital to bear clearly in mind what we have so often (and just recently) said. "We do not intend to predict now what the outcome of the retrial should or may be. All we hold is that the evidence on this trial raises this issue. It is futile to anticipate what the evidence may be on retrial. It is almost certain to be different, and its sufficiency inevitably is a matter for initial determination by the trial Judge applying the principles here laid down but without any artificial effort to match the evidence, bit by bit, against that contained in the present record. Duke v. Sun Oil Co., 5 Cir., 1963, 320 F.2d 853; Smoot v. State Farm Mut. Auto. Ins. Co., 5 Cir., 1962, 299 F.2d 525, 534; Gulf Oil Corp. v. Wright, 5 Cir., 1956, 236 F.2d 46, 53; cf. Carss v. Outboard Marine Corp., 5 Cir., 1958, 252 F.2d 690, 693." Garrett v. American Airlines, Inc., 5 Cir., 1964, 332 F.2d 939 at 944 [1964].

Reversed and remanded.

Answer: No.

Answer: No.

Answer: No.

Answer: No.

Answer: No.

The blowout preventer (BOP) was affixed to the top of the surface casing, just below the derrick floor. Hydraulically operated, it had two sets of "rams." One would completely close off the entire opening of the casing when the drill pipe was not in the hole. The other, used here, was designed to securely close off the space between the casing and the drill stem — called the annulus. In effect, the ram conformed to the drill pipe with the seal being effected by compressible rubber collar-like gaskets.