A fire occurred on Tuesday, March 5, 1968, at about 11:30 p.m. in the Belle Isle Salt Mine while 21 men were working underground.  There were no survivors; 20 died of carbon monoxide poisoning, and one as a result of massive skull fracture.

Although every piece of available evidence was examined in detail during an investigation that required nearly 6 months, neither the cause of the fire nor the point of origin could be definitely established.  It appears that the fire originated in the lower part of the shaft at about, or below, the mining level.

The cause could have been an electrical fault, use of an oxyacetylene torch, or frictional ignition of a belt conveyor, and the evidence does not clearly favor any one of the three possibilities.  Direct property damage was confined to the mine shaft and its equipment.

General Information

The mine is located on the Belle Isle salt dome, along the Gulf Coast in St. Mary Parish, 19 miles southeast of Franklin, Louisiana.  It is one of a group of underground salt mines in similar domes in the area, each of which is operated by a separate company.

The domes are known as Jefferson Island, Avery Island, Weeks Island, Cote Blanche Island, and Belle Isle.  Their surface elevations are not high, but as they were mound-shaped and rose abruptly above the flat marshland, they came to be known as the Five Islands.  The mine, which went into production late in 1962, is owned and operated by Cargill, Incorporated.

The total number of employees was 60, of whom 32 were classified as regular underground employees; some surface employees worked underground intermittently, and staff officials spent much time below.  The mine was operated two shifts a day, and produced an average of 6,400 tons of salt (sodium chloride) daily.

Ordinary maintenance work was done on production shifts.  Although a complete Federal inspection of the mine has never been made, it had been visited by the Bureau of Mines representatives at the request of the mine management on several separate occasions between 1963 and the end of 1967 to examine particular phases of the operation.  The most recent of these visits was made by Bureau mining engineer Arthur M. Evans on August 9, 1967, nearly 7 months before the fire.  His memorandum report, copies of which were mailed to the company on September 13, 1967, included recommendations pertinent to the disaster, that: (a) fire protection should be provided for (among other facilities) shaft stations; and (b) a second shaft should be sunk and connected to the workings.

Mining Methods, Conditions, and Equipment

Geology of the Salt Deposit

Some descriptions of the Gulf Coast salt domes should be helpful in understanding the selection of the location for the shaft, the general plan of mining, and other factors of the Belle Isle operation.  Many salt domes in the Gulf Coast area, including the Belle Isle dome and its neighbors, though all somewhat different in shape, size, and depth below the present surface, when viewed in plan have a generally either cylindrical or elliptical cross section, the axis (or axes) increasing from top to bottom.  The origin of the Gulf Coast domes is from a deep mother bed of salt, thought to be more-or-less continuous over the area.  The weight of overlying sediments is thought to be the force which causes the salt to extrude upward through the overlying sediments, and some salt domes on the continental shelf are said by geologists to be still pushing upward at a rate of perhaps 1 foot every 100 years.

A room-and-pillar system of mining was employed.  Since terrain of Belle Isle is barely above sea level (elel shaft collar 25 feet, 6 inches above mean tide), prevention of surface subsidence is essential, and, therefore pillars were not extracted.

Ventilation and Gases

Flammable, toxic, or noxious gas has not been the determining factor in providing ventilation for this mine.  The mission of ventilation has been to improve the personal comfort and efficiency of employees exposed to a year-round environment of 85-90 degrees F., and high relative humidity.

The shaft was divided by a plywood curtain wall.  The fan drew air down one compartment and directed it into the Air Entry, but the crosscuts were roughly 25 by 25 feet in area and, consequently, difficult to block off.  Bulkheads were built of fine salt, piled as high as feasible, then topped for the last few feet with plywood or plastic brattice material, but the concussion from blasting knocked out these brattices.

As the workings progressed, the effect of concussion near the fan was reduced somewhat, but, up to the time of the fire, effort had been made to close only two crosscuts on each side of the Air Entry.  Beyond that, the air meandered through the workings on both levels until it found its way back to the shaft.

Examinations for methane were not made during normal mine operation.  Gas has been ignited in the kerf on occasion, while undercutting in areas where a shale streak appeared in the salt.  From the characteristic odor and instrument tests, the flammable gas has been taken to be hydrogen sulfide.  Since, however, a flammable concentration of hydrogen sulfide is a highly lethal mixture, whereas mine workers have not experienced even slight hydrogen-sulfide symptoms, and since methane and hydrogen sulfide have been found in association in salt mines, and usually in the vicinity of a shale occurrence, it appears that the principal constituent of the gas ignited in the Belle Isle mine was probably methane.  Analysis of samples collected during recovery operations in the absence of positive ventilation indicated presence of 0.09 percent methane.

Illumination and Smoking

Incandescent lighting was provided generally, but some of the men also used permissible electric cap lamps.  Smoking was prohibited in certain designated areas but was otherwise permitted and practiced.

Fire Hazards and Fire Protection

It has been a common practice for salt mines to use a great deal of wood for many purposes.  Some mines have even the headframe and adjacent buildings of all wood construction.  Basically, wood is used because salt is highly corrosive to common metals.  There is also, however, a belief rather widespread among salt mine people, that when the wood becomes coated and to some extent impregnated with salt, it will not burn.

Apparently, serious fires have not previously occurred in any salt mine in the area, and remnants of timber removed from the Belle Isle shaft proved very difficult to ignite and burn in an ordinary bonfire.  However, many materials considered normally fireproof or fire-resistant will, after preheating, and particularly in the presence of forced-draft air supply, be rapidly and entirely consumed by fire.  Obviously, that is what happened on March 5 at the Belle Isle Mine.

Mine Rescue

Mine rescue teams were not maintained by any of the salt-mining companies in the area.  Since the fire, arrangements have been started by Cargill, Incorporated, and the Bureau of Mines to provide a mine rescue station at Belle Isle Salt Mine, and to train at least two mine rescue teams in recovery and firefighting procedures, as well as to interest neighboring salt-mining companies in doing the same.

Evidence of Activities and Story of Fire

The day's activities on March 5, 1968, were reconstructed from brief entries in the hoist operator's log, supplemented by statements of the second-shift hoistmen and other mine personnel.  The day shift went down at 7:40 a.m.  There were no barges available for loading.  At 3:11 p.m., hoisting and barge-loading of salt was started.

The night shift relieved the day shift underground at about 3:30 p.m., and the day shift crew was hoisted.  Hoisting of salt was resumed and continued for about 6 hours.  For the period 9-10 p.m., the log showed 39 skips hoisted, followed by the remark, "Stop hoist 9:49. No barge."  Nothing unusual had occurred up to this time on this date, and no maintenance work had been done in or about the shaft on either shift up to this time. When out of barges, it was customary to continue mining and processing, and to divert the crushed salt into the underground storage chambers.  The night shift had been working overtime regularly, and on this day it was scheduled to continue crushing until 1:45 a.m.  After the available barges had been loaded, the loading-dock crew went underground to complete their shift, bringing the number underground to 21 men.  These last men went below in the auxiliary cage about 10:25 p.m.

A maintenance crew on this shift was scheduled to lubricate the skips and the skip loader and to make any needed repairs.  This routine was performed at least once a week, and, so far as feasible, when there were no barges to be loaded.  The usual procedure was for the maintenance men to bring their equipment, lubricant, compressed-air grease gun, grease compressor, and arc welder from the underground shop area, using a military-type Jeep.  Oxyacetylene equipment was kept on a hand cart at a safe distance from, but convenient to, the shaft.  Thus, all necessary equipment was at hand at the mining level shaft station.

It is not known definitely what, if anything, besides lubrication, was required to be performed this night, but, after the fire, all the named equipment was found not far from the shaft, and the indications were that all had been used, except possibly, the arc welder.

The lubrication routine was, of course, known to the hoistman.  At approximately 10:15 p.m., the maintenance men signaled for the north skip.  The usual time for the full procedure for one skip was about 15 minutes and, although the hoistman did not record the exact time, all went as usual, and he judged that the elapsed time was about 15 minutes.

Next, the mechanics called for the south skip, and went through the same routine.  The time worked the south skip was estimated also about 15 minutes.  Then they called for the north skip again.  After north skip was sent down, there was a slight delay at the mining level, while, as the hoistman assumed, the mechanics "put whatever stuff they needed" on the cage to work down below.  As explained before, the cage of the north skip was the only one kept in condition to handle men and materials.  They then belled to be lowered to the skip loader, stayed there about 10 minutes, and were hoisted to mining level.  Shortly, they went back down to the skip loader, stayed longer this time, came back to the mining level, and gave the signal releasing the north skip at 11:20 p.m.

This would ordinarily signal completion of the lubrication-repair work, and someone would then telephone the hoistman and report the job finished.  This time, however, they immediately signaled for the south skip again, a most unusual occurrence.  After the south skip had been landed at the mining level station for something like 5 minutes, someone released it by the knocker, and the hoistman received a call by telephone, "Come down with the north side; the shaft is on fire.

The hoistman lowered the north skip, which required about 1 minute.  In his haste, he overshot the landing slightly, but quickly recovered position.  After 2 or 3 minutes, someone gave six or seven rapid signals; then the same man who had called before telephoned again: "The skip is on fire; we can't get on it."  By now, it would appear to have been close to 11:30 p.m.

The hoistman recognized the voice as that of Roy Byron, a topside man.  While the hoist operator was still on the phone, Paul Granger, the underground foreman, cut in and said, "Go to the radio, get some help, get a lot of help."  He repeated the exhortation three times.  Immediately, another voice unrecognized, repeated three or four times, "Pour some water down the shaft."  This was the last communication from the men underground to the surface, and later attempts to call from the surface were fruitless.  There was thus no indication of how the fire started, or exactly which part of the shaft was aflame.

Probable Point of Origin

Since the ignition agency could not be established, the point of origin could not be established definitely, and vice versa.  The evidence seems to point to that part of the shaft below mining level as the most probable point of origin, and several potential sources of ignition were either actually or possibly present in that area.

Summary of Evidence

Although pertinent mine records were made available, and all requested information was provided by company officials and workmen, some of the essential facts are unknown, and, therefore, some of the evidence summarized herewith was arrived at by deduction or conjecture, as will be evident.

• The hoistman on duty had heard no telephone calls between stations underground indicating any trouble prior to the report to him that the shaft was afire.
  
  
• The two telephone calls reporting the fire were terse, but the hoistman quite properly took action as requested, without delaying to ask for details.
  
  
• Of 21 men underground, none survived, and the telephoned reports did not indicate cause of the fire, or point of origin.  It, therefore, becomes necessary to try to determine such by deduction.
  
  
• Dynamite and other supplies had been delivered underground during the shift preceding the shift on which the fire occurred, but they had been removed promptly from the cage, and thence from the shaft area to remote storage points.  No supplies had been taken underground during the second shift.
  
  
• Explosives and blasting agent being used to charge holes at working faces were far from the shaft.
  
  
• Neither high explosives nor the blasting agent was involved in the fire.
  
  
• Fuel oil had been delivered underground on Saturday, March 2, 1968, 3 days before the fire.  The fuel-oil supply line was installed in the shaft in a compartment separate from the skip compartments.  It was open at both top and bottom to avoid build-up of static-pressure head when delivering oil to the portable underground tank, or between deliveries, and the quantity to be delivered was controlled from the surface.  The oil supply line was drained after each delivery of oil.  Although the installation in proximity to shaft timbering was a potential hazard, no evidence implicated the fuel-oil line as a factor in the fire.
  
  
• Gasoline was taken underground promptly upon receipt by boat, in 55-gallon drums, which were removed promptly from the cage and the shaft area.  It was said that two to three drums was the quantity kept on hand underground.  Lubricants were also taken underground in closed container in the cage.  There had been no delivery of gasoline or lubricants for some days before the fire.
  
  
• The fuel-oil storage tank, after refilling, was taken to the haulage area, well over 1,000 feet from the shaft, where it was accessible for refueling equipment.  Gasoline and lubricants were also kept at points a similar distance inby.  None of these stores of flammable materials was involved in the fire.
  
  
• Mining equipment was not ordinarily used near the shaft, and none of the electrical, gasoline-powered, or diesel-powered mining and haulage equipment was involved in the fire.
  
  
• A gasoline-powered Jeep, used by maintenance men to bring equipment to the shaft for lubrication and repair work, was found parked about 350 feet from the shaft, with the air compressor coupled behind it.  These units had not been harmed by the fire.
  
  
• Customary procedure was to park the compressor where it was found, after completing the lubrication routine, and to take the Jeep inby to the shop area.  Since the compressor was still coupled to the Jeep, the indication is that this equipment was hurriedly removed from the shaft arc to get the fuel tanks away from the fire.
  
  
• The main belt, which terminated 50 feet from the shaft, did not ignite, and there was nothing flammable between the shaft and the belt, so that, in summary, the fire was confined to the shaft timbering and equipment.
  
  
• It is necessary first to try to establish the probable point of origin, in order to try to determine the probable cause of the fire.
  
  
• It appears certain that the fire started in the production side of the shaft.  Had it started in the auxillary-cage compartment, which was the downcast-air compartment, men working in the shop area in the main intake-air current would have been aware of smoke in a short time, and before men at the shaft landing, who were in return air.  However, the shop men would not have known the exact location of the fire, whereas, each of the telephone calls was immediately preceded by knocker signals, and the caller, who was the same man each time, said definitely: "----the shaft is on fire," and later "----- the skip is on fire."  Obviously, he was at the telephone on the mining level near the shaft.
  
  
• The shaft was timbered with pine from collar to sump, and a plywood curtain wall was secured to the timbers from collar to just below mining level.  In the service compartment, which extended from the sump to just above mining level, were the ladderway, with landings and handrails at intervals, all of wood construction, and, also, a bucket-elevator, with buckets riveted to a rubber belt, and boxed in with plywood, all being supported by timber sets.
  
  
• The timber in the upper part of the shaft, from the collar to the end of the concrete lining 369 feet, showed the least burning.  Sets near the collar appeared only scorched; farther down the timber was deeply charred but still in place, to about the end of the concrete.  The curtain wall had burned out to about 250 feet from the collar, and partially above that point.
  
  
• The fire damage increased progressively downward, from about -369 feet, the end of the concrete lining, until from -677 feet, roughly midpoint, to mining level, the timber was totally consumed, except for stubs of beams, or plates, protruding from the shaft wall.  (Every set had been hitched into the wall when the shaft was being timbered.)  From mining level to the skip loader, the timbering in the north skip compartment was partially burned but still in place; from the skip loader to the sump screw it was only slightly charred; in the south skip compartment, and in the service compartment adjacent to the south skip compartment, the timbering was almost completely consumed.
  
  
• The telephone and signal systems were both operative when the fire was reported, at which time it is evident that the fire was making rapid headway at about mining level.  Had the fire started above mining level, it would have made still greater headway above by that time, and the relatively small communication cables would, seemingly, have been sufficiently affected to have rendered communication impossible.
  
  
• The maintenance of communications up to that time is an implication that the fire started at, or below, mining level, and there are other such implications.
  
  
• Turning to work in progress before the fire: the hoisting of salt was discontinued at 9:49 p.m., when all barges had been loaded; production and maintenance work were scheduled to continue until 1:45 a.m., the crushed and screened salt to go into storage chambers underground.
  
  
• No maintenance or repair work had been done in or about the shaft on the day of the fire prior to 10:15 p.m., when the mechanics called for the north skip to be brought to mine level.
  
  
• According to testimony from company personnel, fine salt worked into and "gummed up" the grease around the sleeve bearings used on skip and skip-loader mechanisms, so that it was sometimes impossible to force fresh grease through the bearings.  When this occurred, company employees said, it was the practice to heat the bearing housings with an oxyacetylene flame until the pressure from the grease gun could force out the old grease and salt.
  
  
• Reportedly, it was sometimes necessary to heat the metal to cherry red color, and on occasion, the thinned-out grease had ignited.  There was a possibility, too, of igniting nearby wood.  Company policy required mechanics to have fire-extinguishing facilities at hand at all times.  According to testimony, however, a handful of fine salt usually extinguished the blaze.  It was said by a number of persons who had done this work that a fire watch was kept after such heating.
  
  
• On occasion, the mechanics were said to have used the oxyacetylene torch or the arc welder to make repairs to shaft equipment.
  
  
• The north and south skips had been serviced from 10:15 to 10:45 p.m., which was about average time, and which seems to indicate no repair work was done, though it does not entirely preclude heating with the torch.
  
  
• The mechanics next went down to the skip loaders for about 10 minutes, came back to mining level, shortly took the skip back to the skip loaders, stayed about 20 minutes, came back up and released the skip to the hoistman.  This was longer than the usual time for lubrication, if no complications were encountered, but not longer than usual when it became necessary to heat bearings, or to work on the arms which tripped the skip loader chutes.
  
  
• This trip up may have been to get the torch.  This is conjectural, but the torch and tank cart were found about 50 feet from the shaft, with the hoses just doubled back, and not reeled as usual, which seems to indicate that the equipment had been used and was hurriedly removed from proximity to the fire.  However, there was no indication of where the torch might have been used.
  
  
• If it be assumed that a fire got beyond the usual "handful of salt" treatment during the first 10-minute work period at the skip loaders, it might be assumed that the trip up to mine level was to get extinguishers.  But then, it must be assumed further that the mechanics fought the fire for about 20 minutes, failed to get it under control, and retreated to mining level, and still further, assumed that the fire spread 45 feet upward to the mining level in another 10 minutes, to the extent that they were unable to board the cage, and with no positive ventilation in that part of the shaft.
  
  
• The foregoing assumptions are scarcely tenable, because:
  • Extinguishing equipment available in the immediate area was not sufficient to continue fighting a fire for 20 minutes; and,
    
    
  • If the fire had been threatening to get out of control, it appears logical to believe that the mechanics would have called the shop for more extinguishers; and,
    
    
  • If the fire were spreading as assumed, the men could scarcely have endured the heat and smoke for 20 minutes in that portion of the shaft, where air movement was practically nil.
• When the mechanics came up to mining level the second time and released the north skip to the hoistman, this would ordinarily signal completion of the round of maintenance work, and the engineer would also be notified by telephone.  Instead, the south cage was called for again.
  
  
• If it be assumed that a fire was started at the skip loaders during the second, or 20-minute, work period and could not be extinguished, it seems logical to believe that when the mechanics were hoisted to mining level, they would have held the north skip and telephoned the alarm to others in the underground workings, since only the north skip had a usable man cage.
  
  
• It might be assumed that the fire then existed in the south side of the skip-loader pocket, where burning was later found to have been more intense, and that it was intended to use the south skip to get at the fire, but the extinguishers found nearby, some emptied, could not have been collected from inby points in so short a time, and it thus appears that they were collected after the fire was reported.
  
  
• The next presumption is that the mechanics meant to have a second go at servicing something on the south skip, since they had previously serviced it.  If it be assumed that a fire was started by use of the torch during this period of 5 minutes, then it had to spread to the north skip compartment so rapidly, that when the north skip was sent down in answer to the telephone call, a matter of 1 minute from collar to mining level, the north skip could not be approached.
  
  
• The foregoing appears unreasonable.  The north skip was at the mining-level landing for several minutes before it was reported burning.  This had to mean that the plywood siding of the man-cage compartment of the skip was burning, since grease burning on metal parts somewhat remote from the man cage should not have prevented boarding for a 1-minute trip to the surface.
  
  
• The inference is that flames were licking up from below around the cage, and, while the man (men) debated for several minutes whether to chance boarding, the plywood ignited.
  
  
• It can also be hypothesized: that the fire started below mining level, from whatever source, and that the mechanics sent away the north skip and signaled for the south skip, so that the hoistman would take the north skip all the way up and save the cage; that they got the Jeep and the oxygen and acetylene tanks away and then went for extinguishers and help; that Byron, who was a topside man who had been assigned for this shift to duties not far from the shaft, was either notified or was attracted by the commotion; and that, when he came to the shaft and observed the situation, he called for the north skip with the intention to go topside, because he was familiar with the waterline and the fire pump, and he could get such water as was available into the shaft, without taking the hoistman away from the hoist.
  
  
• If the immediately foregoing hypothesis is sound, it would explain the unusual exchange of skips after the customary routine of servicing equipment had been gone through.  It would also explain why Byron, rather than one of the mechanics, reported the fire.  None of the suppositions, however, thus far answers the question of how the fire spread to such proportions as it apparently did, in so short a time.
  
  
• If Byron were indeed trying to get to the waterlines on the surface, and, if he had succeeded, and even assuming that he had encountered no delay in obtaining hose and starting the fire pump, it is improbable that the end result would have been different; the waterline in the belt gallery did not represent adequate firefighting facilities for the shaft.
  
  
• The fire could have burned downward as well as upward, and embers falling from above could have started a secondary fire below; however, if the heart of the fire had been at or above mining level where the velocity of the air current accelerated from about 200 to 1,000 feet per minute, or more, the men would have known that it was futile to fight the fire below mining level, and several fire extinguishers had been used.
  
  
• The fan was installed underground near the shaft and operated blowing.  It could be started and stopped from a control in the shaft area which was readily accessible for at least some time after the fire occurred.  The hoistman observed that the upcast air current had ceased to flow before midnight, possibly by about 11:45 p.m., but it is not known whether the fan had been stopped by those underground, or whether the curtain wall had burned through.  The fact that carbon monoxide was diffused throughout all mine workings explored during recovery operations implies that the fan operated for some time after the curtain wall burned through.
  
  
• The near end of the fan installation was 20 feet from the downcast compartment of the shaft.  A wooden bulkhead surrounded the discharge end of the duct, 50 feet from the downcast.  The existence of some flame and glowing embers at this point, extinguished by the rescue workers some 40 hours after the fire started, indicates that the fan installation was not involved in starting the fire, for, had it been so involved, the bulkhead would have been burned out in much shorter time.  Further, had the fire started at the fan, men in the shop just inby would have been aware of it practically immediately.  Moreover, a fire at this bulkhead would be isolated by solid salt in all directions from any other wood installations.  Thus, it appears certain that this wood structure was ignited sometime later by radiant heat.  The fan and motor appeared to have suffered no appreciable damage, and the V-belts had not burned away.
  
  
• The plywood curtain wall undoubtedly helped to spread the flame from mining level upward; however, it could scarcely have been a factor in the incipient stage of the fire.  None of the potential sources of ignition on the production side of the shaft was close to the curtain wall.  Those which were close, electrical cables and controls mounted on or near the curtain wall in the intake compartment side, are eliminated as sources of ignition, because the evidence is clear that the fire did not originate in the downcast, or so called "air" side, of the shaft.
  
  
• Restriction of fire to the shaft confines possible causes to smoking, an electrical fault, use of torch or arc welder, or frictional ignition involving skip-loader belt or bucket-elevator belt.
  
  
• Smoking was prohibited in the shaft.  It was said that men, on occasion, smoked in the sump, while cleaning up the salt spillage at "the screw."  No such work had been performed on this day.
  
  
• It appears unlikely that mechanics would smoke while engaged in work that was apparently carried on in 1-2-3 fashion, and, in any case, though perhaps not impossible, it appears unlikely that a discarded lighted cigarette or match would ignite wood of the dimensions involved.
  
  
• Electrical cables, motors, and switchgear serving the belt feeder and the conveyor belt were installed in a tunnel below mining level, and to, and at, the screw conveyor in the sump area.  The motor and supply cable for the bucket elevator were located just above mining level.  The lighting circuit extended about the mining level and from top to bottom of the ladderway.  Some of this equipment was destroyed by the fire.  The four circuit breakers in these four motor circuits and one in the lighting circuit were remote from the shaft and were not damaged by the fire.
  
  
• These five circuit breakers were normally in "on" position, leaving the circuits energized up to the individual motor starters or switches.  All were later found in tripped position, indicating that each had been tripped by a fault in its circuit before the primary circuit was opened at the surface about midnight.
  
  
• Had the primary circuit been manually opened at the surface before a fault of low-resistance value occurred in any of the five circuits in question, all five circuit breakers should have remained in normal "on" position, whether the particular unit served by each was operating or idle.  This is inherent in the design of the breakers.
  
  
• Had a fault of low-resistance value occurred in only one of the same five circuits before the primary circuit breaker on the surface was manually opened, the circuit breaker in the faulted circuit should have tripped, but the other four circuit breakers should have remained in the normally "on" position, which could then have been taken to indicate that the fire possibly originated from the faulted circuit.
  
  
• By the usual procedure, the apron feeder and the skip-loader belt would have been stopped when hoisting was suspended, but the sump screw and the bucket-elevating conveyor were usually left running to carry off salt spillage, even when not hoisting, until the end of the second shift.  The incandescent lamps should have been burning.  A fault could, of course, have occurred in an energized circuit before the fire, although the unit served was not operating.  It is, however, not reasonable to believe that a fault occurred in each of the five circuits before the fire occurred.
  
  
• It is possible that an electrical fault in one circuit was the source of ignition, or all five faults which tripped the breakers could have resulted from the effect of the heat from the fire.  Since only about one-half hour elapsed from the report of fire until the surface circuit breaker was deliberately opened, the tripping of all five circuit breakers under discussion appears to be further indication that the fire originated at or below mining level.
  
  
• Of all the circuit breakers on the surface, protecting power and signal lines in the shaft, only one had opened as result of a fault.  This was the circuit breaker in the control circuit of the auxiliary cage.  The heat of the fire presumably created a fault, which caused the hoist to raise the auxiliary cage to within about 300 feet of the collar, when, apparently the fault resulted in the burning out of a transformer in the control circuit, which actuated the circuit breaker.  Engaged otherwise, no one observed this fault-induced activation of the auxiliary hoist, and so, this occurrence sheds no light on when the fire got to the control circuit.  It could have been in the early stages of the fire, or much later, since the auxiliary hoist and its controls were supplied by a surface circuit and were still energized after the underground power supply had been purposely interrupted.
  
  
• Even when power circuits are provided with all conventional protective devices, there is, unfortunately, still the possibility of the occurrence of a high-resistance electrical fault which will not cause the circuit breaker to open, and which, therefore, is the sort of fault which is capable of causing a fire.  For that reason, the possibility of an electrical source of ignition cannot be flatly ruled out, even when, as in this case, no positive evidence of electrical failure is observed.
  
  
• In a tunnel below mining level was the 53-foot belt which carried salt from the surge bin to the top of the skip loader.  The top part of this belt and part of its wooden supporting structure, for about 40 feet from the shaft end, was burned.  Since the back end, including woodwork around the surge bin above did not burn, it appears that the fire could not have started in this tunnel, but that it spread to here from the shaft proper.
  
  
• The torch presents itself as the ready-made culprit.  The indications were that it had been used, but, even so, there was no indication of where, or when, during the hour-long servicing routine.  Granted, that use of the torch to heat grease in proximity to shaft timbers was a decidedly unsafe practice, it is difficult to conceive how a fire started by the torch could have gotten out of hand so quickly.
  
  
• One other possible source of ignition was the bucket elevator.  This unit operated at very slow speed, practically a crawl, which is not highly conducive to frictional heating.  However, assuming a frictional ignition, it would appear that smoke and odor could have been carried up the natural chimney formed by the plywood enclosure, and the fire could have progressed unknown to the mechanics working in the area.  Airflow through the enclosure could have been augmented by the velocity of the mine air current at the head of the enclosure above mining level, a sort of Venturi effect.

Such a fire could have spread very rapidly after it burst out of the enclosure, considering the relatively light lumber used in construction of the facilities in the service compartment.  The extent of destruction in the south skip compartment and in the service compartment, in contrast to the north skip compartment, could be taken as substantiation of this theory.  On the other hand, if the fire first became intense in the south skip compartment, it could have spread more readily to the relatively light wooden installations in the service compartment than to the heavy timber sets in the north compartment. So, again, the evidence is far from conclusive, but a fire originating in the bucket elevator installation might explain the apparently sudden outburst of fire enveloping the shaft at mining level.

Cause of Fire

The cause of the fire could not be determined with certainly.  An electrical source is a possibility, but no positive evidence to sustain such source was observed.  The open flame of the oxyacetylene torch is a distinct possibility; grease fires, though extinguished promptly, reportedly had occurred on occasion, when using the torch to assist lubrication.  Frictional ignition of the rubber belt of the bucket elevating conveyor cannot be ignored, particularly in view of its plywood enclosure and the contiguous ladderways and timbers.

Conclusions

Although the cause of the fire could not be determined, the investigation following the disaster revealed three factors that must, in the Bureau's opinion, be considered as contributing in a major degree to the loss of the miners' lives.  Those factors are:

• The absence of adequate fire prevention measures in a shaft that incorporated a great deal of flammable light timber and plywood in its structure and its facilities.
  
  
• The inadequacy of firefighting facilities at and in the shaft.
  
  
• The lack of a separate shaft, which could have provided the trapped men with another way out of the mine.

It must be noted that these inadequacies had been called to the attention of the company management nearly 6 months before the disastrous fire occurred.  Had a second shaft been begun then, it could not have been completed in time to have had any effect on the outcome of the fire.  The shaft had not, however, been started at the time the disaster occurred.

Source:
	Historical Summary of Mine Disasters in the United States - Volume III