CHRISTENSON, District Judge.
Plaintiff’s equipment being used in the drilling of a gas well was damaged by fire while it was insured by the defendant company against the risk of fire in general but not including “any * * * damage * * * caused by or incident to blowout * * * or any fire loss or damage resulting therefrom.” A jury was waived and the Court has heard the evidence and has considered counsel’s arguments, both oral and written.
The determinative issues of fact, as reserved in the pre-trial order are whether there was a blowout affecting plaintiff’s well and, if so, whether the damage of which plaintiff complains was caused by, or incident to, such blowout, or was fire damage resulting therefrom. If the answer is in the affirmative, plaintiff’s loss was not covered by the policy and the defendant should prevail here; if in the negative, the plaintiff is entitled to recover. It is agreed that the burden of proof rests upon the defendant to establish that the risk was excluded by the policy.
The Court finds the following facts according to its conception of the preponderance of the evidence:
1. The plaintiff now is, and during all material times has been, a corporation incorporated under the laws of the State of Utah, and maintaining its office and principal place of business in Salt Lake County, Utah.
2. The defendant now is, and during all times material has been, a corporation incorporated under the laws of the State of Connecticut, duly authorized to write insurance of the nature involved in this ease and doing such business in the State of Utah as a foreign corporation.
[832]*8323. The amount in controversy in this case exceeds the sum of $3,000, exclusive of interest and costs.
4. On April 2nd, 1953, and continuing until after the events hereinafter mentioned occurred, defendant insured by written contract of insurance certain oil or gas well drilling equipment belonging to the plaintiif. On the 15th day of December 1954, said equipment, so insured, was located within Rio Blanco County, Colorado, on a well site known as “Brennan No. 1.”
5. That among the provisions of the policy of insurance covering said equipment were the following:
“4. While in buildings and while elsewhere at land locations, this policy insures against loss or damage by (except as herein provided):
“(a) Fire and lightning;
“(b) Tornado, cyclone and windstorm ;
“(c) Strike, riot and civil commotion;
“(d) Explosion (except as hereinafter excluded).
“5. This policy does not insure: *****
“(g) Any loss or damage to property caused by or incident to blowout or cratering of an oil or gas well or any fire loss or damage resulting therefrom.”
6. Within the time specified in the contract of insurance, due proof of loss was made and the plaintiff performed all other conditions on its part to be done as required by the insurance contract.
7. On December 15, 1954, oil well drilling equipment belonging to the plaintiff and then being used in drilling the Brennan No. 1 well was lost or damaged by fire to the extent of $42,903.71.
8. That said fire occurred under the following circumstances:
(a) The plaintiif began drilling said well on or about October 20, 1954. Ten and three-fourths-inch casing was cemented at 265 feet with 203 sacks on October twenty-fifth. Seven-inch casing was cemented at 2,025 feet with 270 sacks on November 4, 1954. Drilling below the seven-inch casing proceeded with a mud analysis truck on location. The mud analysis log indicated a gas show at about 2,170 feet and cores were cut from 2,184 to 2,198 feet, which was in the Douglas Creek formation, a recognized gas bearing formation. The formation at this location was of low permeability and averaged less than 0.22 millidarcies according to an analysis of the cores from the well. Halliburton drill stem test was made November 10, 1954 at the interval 2,180 to 2,198 feet using 5^-inch bottom choke and one-inch top choke. Gas reached the surface in 4 minutes with a good blow but there was no actual measurement of surface pressure or gas rate. The rate was, in any event, small. The tool was open 37 minutes. The initial bottom hole flowing pressure was 91 pounds per square inch which reduced to 49 pounds during the test. The tool was closed and the bottom hole pressure built up to 285 pounds per square inch in 15 minutes and was still building up when the test ended.
(b) The well was drilled to a total depth of 4,889 feet which was reached December 4, 1954, with no indication of any additional oil or gas zone below 2,200 feet. The well was plugged back to 2,379 feet on December 7, 1954, and a 5-inch liner cemented on the bottom at 2,379 feet with the top of the liner at 1,956 feet. Tubing was run in the hole and the mud in the hole was displaced with a mixture of Rangely crude and Diesel oil and the liner was perforated December 13, 1954 from 2,174 to 2,180 and 2,184 to 2,188 feet with four bullets and four jets per foot. All of the oil was then swabbed from the hole to test the flow capacity. A small rate of gas flow was observed at the surface, but it was too limited a rate to register on the gauge then being used. The gas was turned to flare and burned with a flame 2 to 3 feet high, indicating a yield of approximately 25,000 cubic feet per day, with no oil showing. The well was then closed and the pressure built up to 320 [833]*833pounds per square inch on a gauge of undisclosed type and accuracy and after a shut-in period of less than 24 hours.
(c) About 95 barrels of oil was pumped back in to fill the hole. The hydrostatic head in the well thereupon was approximately 773 pounds per square inch at the face of the formation. In. preparation for hydraulic fracturing, 2% inch hollow tubing was run into the hole to a depth of 1,860 feet with a 6% inch packer on the end which had not been set and which moved freely in the hole. During the running of the last five or six stands of this tubing of approximately 90 feet each, prior to reaching 1,860 feet as aforesaid, small quantities of oil flowed over the top of the tubing as it was being lowered in the hole. The blowout preventers were left open. The overflow was not considered by the operators to be interfering with the progress of the work. The first overflow occurred when the top of the tubing hanging in the elevators had been lowered to approximately 15 feet above the derrick floor. As each successive stand was lowered the overflow occurred at increasingly greater heights. When the final stand was being screwed in place, oil was coming at the top of the tubing at the elevators approximately 96 feet above the derrick floor and 108 feet above the ground. This condition was the last any eye-witnesses had observed at, or immediately before, the start of the fire.
(d) Within moments after the final stand was screwed in, without audible explosion, a fire suddenly started at, or above, the end of the tubing 96 feet above the derrick floor. Flames immediately shot up more than 30 feet above the crown block, or more than 60 feet above the top of the tubing. The time the fire began was between 4 and 5 o’clock a. m. of December 15, 1954. The weather was then cold and clear and there was a breeze blowing. The fire continued for approximately 30 minutes out of control with sufficient heat to buckle the derrick and heat the crown block red. Burning oil dropped below the top of the tubing and caused damage to equipment below. The fire at the top of the tubing gradually diminished in intensity until it was burning a flame approximately 1 foot high above the tubing at the end of about half an hour. In excess of 15 barrels of oil were discharged through said tubing during and immediately preceding the fire, but the packer and other tools were not cleaned out of the hole.
(e) Seven months later, the remaining oil was swabbed from the hole, at which time there were no pressure indications except that about 5 barrels of oil ran out when the hole was opened up on the latter occasion. After the well was acidized in July, 1955, it produced up to 800,-000 to 1,000,000 cubic feet of gas per day.
9. The evidence does not permit any finding of the exact agency or cause which ignited the fire. The Court believes from the evidence, and so finds, however, that the fire started from a spark or other source of ignition at, or above, the top of the tubing as it came in contact with gas then being discharged from the top of the tubing as a result of an excess of bottom hole pressure above the hydrostatic pressure of the oil in the hole.
From the findings of fact made thus far, it seems now necessary to digress for a consideration of certain legal concepts and factual analyses. Upon these depend the making — perhaps the necessity — of certain other ultimate findings to dispose of the question of liability.
If, in effect, says one side of this controversy, a buildup of bottom hole pressure caused the discharge of gas and oil from the tubing without which the fire would not have occurred, that should be an end to it; the fire was the result of, or incident to, a blowout. Moreover, the manifestation of the fire was violent, sudden and unexpected. It is not a question, says the other, of whether the latter is true, but whether the discharge of gas and oil from the hole as observed prior to the fire involved these elements and to such an extent that the hole was completely “cleaned out” of tools as well as fluid; that there being no competent [834]*834evidence from which the Court can find which of several possible causes ignited the fire, the party having the burden of showing causal connection between the blowout, if any, and the fire, must fail, and in any event, doubts should be resolved against the insurer in view of the uncertainties of the policy. Both sides rely heavily upon expert testimony that the facts specified above either did, or did not, indicate a blowout; and it was suggested by one that the Court is unduly concerning itself with detail since cases of this nature should be broadly determined on the basis of the more convincing opinions.
If the experts’ conclusions were looked to alone, there might be enough to permit, and perhaps to sustain, a decision either way. However, the process would not be a satisfying one. Typically, the expert opinion was in sharp conflict. Both sides presented men of broad training and experience who maintained their respective opinions with similar authority and resolution, and the differences in these respects were not of themselves significant. The expert testimony alone does not contain the answer.
I have found it necessary to attempt a somewhat detailed analysis of the law and the operation and effect of established facts in the light of it. The case is not an easy one but this process has satisfied me of what the solution should be as founded upon the greater weight of the evidence and as compelled by reason. A recordation of this process here will at least serve the purpose of permitting an effective review of any points on which I may have erred at the expense of substantial justice. When a case reaches the appellate court the question often is not what the lower court should have decided but what it could have decided without being reversed, which may be a different thing. As a result of specific, rather than broad, decision on the trial level here it may be, too, that when this case finally has been passed upon insured generally will be encouraged to more fully appraise, and perhaps reject, policies wherein the exclusion is broader than they wish and the insurers to clearly broaden the language of the exclusion if the term “blowout” connotes less than the corresponding premium rate justifies.
Merely because experts or attorneys differ as to the meaning of a term does not demonstrate that it is an ambiguous one to be resolved favorably to an insured. Thomas v. Continental Casualty Co., 10 Cir., 1955, 225 F.2d 798; Commercial Standard Fire & Marine Co. v. Beard Well Servicing Co., 10 Cir., 1954, 210 F.2d 560. The term in question is not incapable of reasonably precise definition. There will always be areas of uncertainty along the perimeter of a meaning even after it has been compressed to the fullest extent practicable. It is only if the facts cast this case within such a narrow area that the pertinent rule of construction will need further consideration.
Plaintiff relies strongly upon the definition of a blowout contained in Central Manufacturers’ Mut. Ins. Co. v. Elliott, 10 Cir., 1949,177 F.2d 1011,1012, wherein Judge Huxman, speaking for a unanimous court, states:
“The term ‘blow-out’ as used in oil operations has a technical meaning. It is generally defined as a condition in which a well builds up a sufficient gas pressure at the bottom of the hole and causes a rather sudden, forceful eruption or explosion which cleans out the well and causes it to go out of control.”
Since the “cleaning out” of a well, argues plaintiff, means the cleaning out of the tools, among other things, there can be no blowout if the tools in the hole are not blown out. I do not believe Judge Huxman’s definition dictates any such construction. If the tools are “set” so that the gas cannot escape without blowing them out, obviously there could be no blowout unless the tools are cleaned out, or at least unless the controls are released in some manner. In this I am not unmindful of the control maintained of the wells involved in Georgia Home Ins. Co. v. Means, 5 Cir., 1951, 186 F.2d 783 and Fidelity-Phenix Fire Insurance [835]*835Co. of New York v. Dyer, 5 Cir., 1955, 220 F.2d 697, but there were involved special policy definitions which are not present here. With the exception of the one element involved in the foregoing contention of plaintiff, both sides pretty well agree upon ths meaning of the term “blowout” in the abstract. Such accepted meaning was not inconsistent with Judge Huxman’s basic definition, which I use as my guide. It is necessary on this phase, however, to directly apply the general definition to the special elements in this case, and to arrive at a more particular interpretation with these elements in mind.
If the situation of the tools permits the blowing-out of a well by the cleaning out of oil, mud or other liquid designed to control it, so that the well goes out of control violently, suddenly and unexpectedly by the buildup of bottom hole pressure, there seems no reason why these circumstances could not involve a blowout despite the continued presence of the tools in place, or despite the resumption of control at a later time. However, if a mere flow or welling-over of a liquid is present, with no substantial violence or suddenness and with no eruption of such a nature as to clean out that fluid which is between the source of the pressure and the outlet and which is blocking its release, it seems plain that there would be no blowout. This was inferentially recognized by defendant’s experts. One of them called such a condition short of a blowout an “incipient blowout”. It has been otherwise referred to as a “kick”. Fidelity-Phenix Fire Insurance Co. of New York v. Dyer, supra.
Not every escape of fluid from a well, even though to some extent forceful, unexpected, uncontrolled and undesired, would be a blowout. As a blowout develops from an incipient state to a blowout in the proper sense of the term the bottom hole pressure causes sufficient fluid to flow out of, or to erupt from, the hole to further increase its imbalance with the hydrostatic head. Once this process begins, the divergence between the two ordinarily increases, with consequent increase in the violence of the eruption until the excessive pressure is spent. The more liquid discharged, the greater the disparity between these two forces may become. This seems the reason — consistent with Judge Huxman’s definition — that a blowout ordinarily does not end of its own accord with only part of the hole cleaned out.
Hence, a “blowout” in view of the facts in this particular case, means a condition, incompatible with normal operations, in which a well builds up sufficient bottom hole pressure to cause a sudden, forceful and undesired eruption, cleaning out any controls which are retaining the pressure in the hole, whether liquid, valves or tools, so that the pressure is violently released into the air. If the excessive pressure is not so released, there is no blowout, although the effect may be a cratering of the well within another arm of the exclusion, or other condition. The eruption need not be of the extreme violence exemplified in Green v. General Petroleum Corp., 205 Cal. 328, 270 P. 952, 60 A.L.R. 475, for the exclusion is not made dependent upon the degree of violence or the extent of the damage resulting from it. But the discharge of fluid must constitute an eruption as distinguished from a mere flow or welling-over. The extent of efforts to control such eruption is not determinative so long as the occurrence is unexpected and undesired; nor is it material that efforts or means for control are short of those dictated by the most effective procedures or even due care so long as the condition is not wilfully or wantonly created.
I am, therefore, of the opinion that the welling-over of the liquid as last observed before the fire was not a blowout. It was not sudden nor unexpected because it had been continuing, although with gradually increasing pressure from below, while four other stands were being lowered into the hole. It was not violent, and, while it was undesired, it was not incompatible with normal operations. An “incipient blowout” is not a “blowout” within the terms of the policy. If all escapes of fluid from a well were [836]*836meant to be covered by the exclusion, much more apt language than a “blowout” could be used in an insurance policy to cover it. “Any damage caused by or incident to the escape of gas or other fluid from the well into the air” might be sufficiently broad. Certainly, the term “blowout” would not be. Hence, recovery could be had for damages by fire resulting from an incipient blowout or other conditions falling short of a blowout. This is made clear in Elliott, supra, where the following comment was made by Judge Huxman, for the court, in sustaining the lower court’s judgment on the facts in that case, 177 F.2d at page 1012:
“There was a great deal of testimony warranting the conclusion that there was no sudden, violent eruption which cleaned out the hole, and that the well was not out of control prior to the explosion or fire, but rather that there was a gradual escape of gas through the mud and water in the well that settled around the well which was ignited by a spark when the engine was started. * -* *»
If, however, between the time the fire actually started and the last observed condition preceding it, an eruption took place as the result of the continued buildup and sudden release of bottom hole pressure, or if the fire started simultaneously with such eruption and as a result thereof, it might properly be said that the fire was the result of a blowout.
This case, therefore, turns upon a rather narrow question of fact. It cannot be fully resolved upon the basis of direct testimony, and the expert testimony is not conclusive either way. The physical facts and matters of common knowledge must be looked to primarily for the answer. There seems no reason why circumstantial evidence cannot be considered in this as in other cases. The Rocona v. Guy F. Atkinson Co., 9 Cir., 1949, 173 F.2d 661; Auto Transport v. Potter, 8 Cir., 1952, 197 F.2d 907; The Wenona, 19 Wall. 41, 22 L.Ed. 52. And physical facts within common knowledge and reason may be sufficient to establish the truth of a matter, especially as against inclusive evidence or speculation to the contrary. Baltimore & O. R. Co. v. O’Neill, 6 Cir., 1911, 186 F. 13; Rosenberg v. Baum, 10 Cir., 1946, 153 F.2d 10.
I believe that a blowout in the proper sense of the term did occur immediately preceding, or simultaneously with, the start of the fire.
(a) The manifestations of the fire as established by the evidence without dispute were inconsistent with any theory that it could have been started in, or could have been fed by, the flow of oil from the tube as last observed. What first called attention to any different or unusual situation after the last stand was connected was the sudden shooting up of the flames and the burning oil falling through the air. The flash point of oil was shown to be 100° F. Even the application of a match would not have ignited the oil itself. Moderate quantities of gas moving through the oil into the air would have been dispersed immediately by the breeze, and in any event, if ignited would have burned moderately, as when the flare test was theretofore made. The gas could not have burned down inside the tube without oxygen. Even though the oil had been ignited, the burning surface at the top of the tube was exceedingly limited, and the fire itself could not have been the cause of any substantial increase in the imbalance between the gas pressure and the hydrostatic head.
(b) The occurrence was consistent only with the theory that there was a blowout. A blowout would have rendered the fire most likely. The violent discharge of large quantities of gas and oil into the air would explain the suddenness of the fire, its violence, its continuation at high levels to such an extent that the crown block, high above the top of the tube, was red hot and buckled by the heat, and the showering of burning oil from above the tubing. The fact that after about thirty minutes the flame had decreased to the height of approximately a foot indicated that the bottom hole pressure progressively had spent itself [837]*837from the eruption. I have been unable to explain the action of the fire on any other theory.
(c) The check of the well following the fire indicated that a substantial quantity of oil had been lost from the hole. Only a relatively small amount would have welled over the top of the tube prior to the fire. As nearly as can be determined now the loss, in addition to this amount, was the oil which was cleaned out of the tubing, and the casing below the packer, as a result of the escape of the excessive pressure. This oil was the only effective control upon the escape of such pressure. True, and plaintiff makes a great point of this, the oil in the annular space between the tubing and the casing was not discharged, and according to the estimate of the witness Rooks, about 80 barrels of oil remained in the hole after the fire. This would put the amount of oil cleaned from the hole at less than 15 barrels. However, Rooks said that the remaining oil was not measured, and that his reference to its quantity was a mere guess. Moreover, the evidence shows that the “packer” at the bottom of the tubing, while moving freely within the casing, left a constricted opening around its sides. The bulk of the oil was in the annular space which was protected at least to some extent from the immediate effect of bottom hole pressure by the packer. This must have been so, or the oil would have been discharged from the annular space into the mud pits, or above, when the pressure became great enough to raise the oil more than 90 feet in the tubing above.
One of the plaintiff’s theses was that the rise of the oil in the stands was not from bottom hole pressure but was the result solely of the lowering of the tubing into the oil in the hole, with much the same effect “as if a straw were lowered into a seven-up bottle where fluid came out of the end of the straw.” It cannot be gainsaid that there would be a displacement of oil by the metal in the tube, but its effect could not be to cause gas and oil to erupt violently into the air after the lowering process had stopped. Moreover, this theory recognizes the restriction the packer would have furnished to the passage of oil from the bottom of the hole into the annular space above the packer, or it encounters the same paradox which plaintiff says the defendant’s theory involved: the elevation of the oil in the tubing but not in the annular space. Suffice it to say that the gas and oil did rise in and overflow the tubing and did thereafter violently erupt; and to the extent that they did not also erupt from the annular space into the mud tanks, or above, it is obvious that they were blocked by the packer.
(d) The last measured test of the well before the oil was put in indicated a continued buildup of pressure, however moderate. When the gas was put to flare before the fire it burned with a flame two or three feet high, indicating a yield of about 25,000 cubic feet of gas a day. Immediately upon the start of the fire flames shot up more than 60 feet above the top of the tubing. After an interval of about a half hour, the flame had reduced to about a -foot in height. After acidizing, the well produced about 1,000,-000 cubic feet of gas per day. The nature and sequence of these events seem persuasive of the conclusion reached.
(e) All of the circumstances supporting defendant’s theory of blowout had to be drawn from employees of the plaintiff by the defendant which had no representatives at the scene. This is not to say that their testimony was not substantially accurate as far as it went, but it at least may be supposed that all of the facts favorable to the plaintiff’s theory were made available. The Court, under such circumstances, should not impose upon the defendant — indeed may not — the unrealistic burden of answering every possible question, as distinguished from sustaining its position by the greater weight of evidence and reason.
(f) The expert testimony favorable to defendant’s position is believed more adequately supported by fact and reason and therefore more convincing.
It having been determined that a blowout occurred, it remains to be seen [838]*838whether this blowout invoked the operation of the exclusion clause of the policy in question. The pertinent terms of the exclusion, omitting considerations not vital to the particular question are “any * * * damage * * * caused by or incident to blowout * ■ * * or any fire damage resulting therefrom.” Fire damage resulting from what? From damage caused by, or incident to blowout, or from blowout itself? It seems obvious that it is the latter. There is expressly excluded not only damage caused by, but damage incident to a blowout, in the event that such former phrase might be considered not to comprehend the latter. But this refers to the damage directly caused by, or incident to blowout, as where the force of the eruption blows out the tools or directly affects the superstructure or otherwise directly causes damage. For'an example of the latter type of damage, see FidelityPhenix Fire Ins. Co. v. Dyer, supra. As to fire damage, it is not sufficient that it be merely incidental to a blowout. It must proximately result from blowout to be excluded. The language certainly means this, and if it does not, it at least is so ambiguous as to call for resolution in meaning against the insurer on this point. Such a construction is supported by the statement in Feeney & Meyers v. Empire State Insurance Co., 10 Cir., 1955, 228 F.2d 770, 771, where it was recognized with reference to similar language in the policy involved in Elliott, supra, that the only question there was “whether there was a blowout, and if so, whether it caused the fire.”
Did the blowout cause the fire? That the Court cannot say which of several possible agencies actually ignited the fire is not conclusive on the. question of causation between the blowout and the fire. The cases cited by plaintiff to the effect that when it cannot be determined which of several circumstances causes a given occurrence, a finding of proximate cause cannot be sustained, are not in point. Typical is Bean v. Independent Torpedo Co., 8 Cir., 1925, 4 F.2d 504. There it was held in an action for the death of an employee where there was no evidence of the cause of an explosion which might have been due to any one of several causes, that the jury could not be permitted to guess at the cause.
Here, irrespective of possible precipitating causes, the question is whether the blowout was a proximate cause. It would not matter whether the igniting agent was a spark, a match, static electricity, or, as seems most likely, the result of the striking of solid material discharge from the hole upon the superstructure of the rig. Nor would it matter whether the actual igniting was due to negligence or similar circumstance.
The answer on the issue of causation must be sought in the light of the particular problem which is before us. This is not whether a spark or some act of the plaintiff caused the fire, but whether, despite the exact agency of ignition, the fire resulted from the blowout. It is apparent that it did. The fire resulted also from a spark or other igniting agency; but one cause does not necessarily exclude the other as an efficient cause for the purposes of a specific case. Parkinson v. California Co., 10 Cir., 1956, 233 F.2d 432. With reference to a particular policy there may be a distinction drawn involving proximate cause as between insured and insurer, and for other purposes. See American Insurance Co. of City of Newark, New Jersey v. Keane, D.C.Cir., 1956, 233 F.2d 354, 358. The case of Lanasa Fruit Steamship & Importing Co. v. Universal Ins. Co., 302 U. S. 556, 58 S.Ct. 371, 82 L.Ed. 422, also furnishes the answer on the basis of a related principle.
In view of the foregoing considerations I make further findings of ultimate fact as follows:
10. That at the time of, or immediately preceding, the said fire a blowout occurred within the contemplation of the policy of insurance involved here.
11. That said fire was proximately caused by said blowout and resulted therefrom.
Accordingly, it is concluded that the risk of said fire was excluded by the [839]*839terms of the insurance policy and that plaintiff is not entitled to recovery thereon.
This memorandum decision may be deemed the Court’s findings of fact and conclusions of law, and the clerk, in accordance therewith, is directed to enter judgment that the plaintiff take nothing; that its action be dismissed on the merits; and that the defendant recover of the plaintiff its costs.