National Academies Press: OpenBook

Guidelines for the Investigation of Grain Dust Explosions (1983)

Chapter: Appendix: Investigations

« Previous: Chapter 3: Summary of On-Site Investigations
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 17
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 18
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 19
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 20
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 21
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 22
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 23
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 24
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 25
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 26
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 27
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 28
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 29
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 30
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 31
Suggested Citation:"Appendix: Investigations." National Research Council. 1983. Guidelines for the Investigation of Grain Dust Explosions. Washington, DC: The National Academies Press. doi: 10.17226/10954.
×
Page 32

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

APPENDIX INVESTIGATIONS During the period November 1978 to July 1981, the sites of a number of grain elevator explosions were visited by members of the Explosion Investigation Subpanel of the Panel on Causes and Prevention of Grain Elevator Explosions.- The elevators ranged in size from a small country elevator with a capacity of approximately 20,000 bushels to a large export elevator with a capacity of approximately 6,000,000 bushels. In all cases, excellent technical information on grain dust explosions was obtained. If one includes two recent European investigations--the Roland Flour Mill in Bremen and Euro silo in Ghent--this information can probably be considered the most reliable contemporary data in existence. The six incidents described here are typical. Information from each accident is presented separately in the form of an incident report. In the first section, Observations, factual information collected at the site of the accident is given. In the second section, Scenario, a plausible chain of events leading up to the accident is presented. Some of this material is speculative. In the-third section, Conclusions, information relating to grain elevator safety and the successful conduct of investigations is discussed. Incident No. 1 An explosion occurred at a concrete inland terminal elevator of medium size. There were truck dump and loading facilities and rail dump and rail loading facilities. A large headhouse was located between, but separated from two silo complexes. There were three bucket elevators within the headhouse having combination steel and concrete double legwells. One truck dump was adjacent to the headhouse and not enclosed, another was located a short distance away and enclosed. The rail deep and loading areas were alongside the headhouse on two lines; the inside line served the dump pits. There was a large drier between the headhouse and one silo complex with two bucket elevators. A machine shop was located on the opposite side of the headhouse. The office was located a moderate distance from the main elevator accident. There was a primitive dust control system. The level of housekeeping, as judged qualitatively by the subpanel, was not good. Observations The explosion occurred in mid-afternoon. 1 17 Severe structural damage was

18 done to almost all of the bins in the headhouse and moderate damage was done to most of the headhouse structure. The tops of almost all of the headhouse bins had blown off, essentially destroying the top of the bin floor and the top of the cleaner floor. In addition, some of these bins around the edge of the headhouse caused the failure of the outside wall. In the headhouse structure above the bin floor there were a large number of casement type windows. All of these windows were destroyed but there was little damage to the walls themselves. Fireballs had propagated through the basement, work floor, bin floor, and scale floors. Severe damage occurred to all elevator legs. Where the legs passed through the house bins in concrete legwells in some areas the concrete had completely shattered. This was especially true on the gallery floor where one of the legs vented. On other floors above and below the bins many of the the metal leg casings were split wide open. The one elevator leg showed severe fire damage with much of the belt and many plastic buckets burned. The exterior wee and dry drier legs also showed moderate explosion damage. Adjacent to the work floor the rail loading and dumping area showed fire damage but only alight blast damage. The two truck dumps showed evidence of fire and blast damage. The dust collection systems adjacent to the headhouse showed slight explosion damage. In the one silo complex severe explosion damage occurred to the catwalk connecting it to the headhouse, the gallery at the far end of the tunnel, and a small group of silos centered around an airshaft approximately one-third of the way along the gallery. This happened to be the location where the belt tripper was parted. Between the headhouse and this location, gallery sidewalls and the roof had been displaced and the windows were destroyed. At the location of this airshaft the gallery walls and roof had been completely destroyed. Beyond this point the explosion damage to the gallery was still significant but not as severe. At the location of the airshaft the grain in the surrounding tanks was on fire. The interior concrete bin walls had been extensively shattered leaving in many places only the reinforcing rod. The tripper rails above this area had been bent straight up. Concrete fragments from this area of the elevator had been thrown several hundred feet into the adjoining rail yard. In the tunnel under this silo complex only the portion from the airshaft to the far end showed any fire or explosion damage. An extremely strong blast wave had propagated down this tunnel shearing of f the grain Spouts at the level of the bin bottoms and completely destroying the end of the tunnel. The vertical slab had been moved several inches back into the earth. The other silo complex showed practically no explosion damage with the exception of an empty bin near the far end, which had exploded and caused moderate damage at its bottom. Damage to the tunnel was slight and a fireball had propagated down the gallery from the headhouse. Some damage from flying debris was done to the office building and cars parked near the elevator. At the time of the explosion milo had been unloaded from rail cars and corn was being loaded into rail cars. The leg that had been carrying the milo was running empty and another leg was carrying the corn. Both of the basement belts were running and neither gallery belt was being used. The first evidence of an explosion came from an employee in the basement near the north tunnel. He stated that he heard a "pop" and saw a f ireball coming out of the boot of the leg that had been running empty. An employee on the scale f loor f elt

19 the explosion and sought shelter under a desk in the scale shanty. An employee standing outside the office and looking at the basement of the headhouse saw dust coming out of the basement windows and heard five explosions. An employee on top of the railroad car saw a fireball come out of the loading spout and was knocked to the ground. He thought that he heard three explosions . Concre te f ragment s came through the roof of the of f ice building but did not ~ trike any of the occupant s . An individual here described the explosion as sounding like a train wreck or the bumping of cars in a railroad yard. At this point the plant manager attempted to call the fire department but the telephone was inoperative. Personnel from surrounding industry called the fire department, who extinguished the fires in the headhouse but not in the grain tanks near the airshaft. After the explosion, regulatory agency personnel found extensive tramp metal in two of the elevator boot pits. Elevator employees unloaded the hopper care that had been filled near -the time of the explosion and found a bindicator (bin indicator) and a portion of electrical conduit. The bindicator had been located in the loadout bin. The fuse associated with this bindicator was blown and the wiring and conduit remaining in the bin showed evidence of arcing. Scenario A low level bindicator was installed in the loadout surge bin, which was between the scale and the car. Near the bottom of the bin a bindicator was installed inside the bin wall cantilevered on an electrical conduit elbow. The bindicator weighed perhaps 10 pounds, was approximately 10 inches in diameter, and had a depth of approximately 6 inches. The-face was a rubber metal diaphragm, which made mechanical contact with a microswitch. Appropriate installation of this bindicator would have made the diaphragm flush with the bin wall. A load of grain that was dumped from the-scale impacted upon the bindicator causing it to break loose from the conduit. - This left exposed at the conduit the bindicator circuit wires with a live potential of 110 volts. After several additional grain dumps from the scale into the surge bin one of the dumps impacted upon the broken conduit and exposed wires. This caused the wires to contact each other and the conduit, producing an electrical arc. As the scale was continuing to dump the surge that was ignited by the electrical arc, an explosion occurred in the loadout bin and propagated into the adjoining house bins and the bin floor area through the destruction of the bin walls and tops. The explosion was able to go out the loadout spout to the car. The explosion also traveled up the spout into the scale and then through the common dust collection system into the bucket elevators. The legs blew out at various levels in the headhouse. The leg blowout on the gallery floor allowed the explosion to propagate in one direction toward the set of tanks where the airshaft was located. Progress of the advancing explosion down the gallery was impeded by the tripper, and the advancing airflow went down the ventilation shaft. This airflow in a vertical shaft would be ideal for producing a well-mixed dust-air mixture. The flame front then arrived as it too was deflected by the tripper down this shaft. A very rapid rate of combustion occurred. The combustion wave then came back up the shaft and returned toward the

' 20 headhouse and went down the shaft into the tunnel where it went out the far end. In the gallery a flame front from the exploded leg casing traveled into the other silo complex. Conditions there were not as favorable for , propagation and only a low velocity flame front traveled near the floor. However, at the end of the closed gallery it was reflected into the top of an empty tank. The dust inside the tank was disturbed and an explosion occurred. Conclusions . Regulatory agencies have listed the tramp metal in the boot pit as the cause of this explosion. All reasonable efforts should be made to keep this foreign material out of the elevator equipment by the use of properly sized gratings on all dumps and the use of magnets on bucket elevator feeds. Regular cleaning of the boot pit will catch any foreign material that escaped the other collection system. However, there seems to be little information relating to the detailed mechanism of the heating and ignition of grain dust by tramp metal. ' The ignition source considered by investigators from the subpanel to be more likely is the failed Vindicator. Bindicators should, of course, be installed in the specified fashion. Moreover, only' low voltage electrical circuits should be used ' in Vindicator circuits to lessen the hazard caused by arcing. Guidelines for design of circuits that are incapable of igniting grain dust are contained in NFPA 493, "Intrinsically Safe Apparatus for Use in Division 1 Hazardous Locations." Other possible ignition sources that were rejected as causes because they were not at the apparent point-of-origin were rubbing of the gallery belt on the frame and the dragging of rail cars through the car dump using a winch. The rubbing had begun to cut the frame and there is little doubt that the metal was red hot. This could have led to a smoldering fire. Friction sparks that could be causes by the dragging of the cars could have fallen into the dump pit. It may have been possible to have detected the loss of the Vindicator. If there had been a grating on the hopper car fill the broken Vindicator would have been caught at this point. If a truck had been filling, presumably the driver would have seen it. Also, if the Vindicator signal had been continually monitored, the loss of signal would have been immediately noticed. A minicomputer is more reliable than a human in this respect . Once the explosion was initiated, it could propagate in several directions. A Mayo spout was used on the bin floor. Once the explosion reached thi 8 location it could enter every headhouse bin. Also, the dust collection systems on the scale and the legs were interconnected, allowing the explosion to propagate into the legs. European design practice is to use several smaller dust control systems. The one leg was running empty and there was probably a heavy suspended dust concentration. The legs

21 effectively propagated the explosion downward. The headhouse had an unusually large window area for its volume; it was quite well vented and there was little structural damage. Most structural damage was done from the explosion of headhouse bins (garner bins). Concrete fragments were dispersed to rather distant places. Some of them impacted on the office building. The commonly mentioned rule of thumb that a distance from the elevator equal to the height of the headhouse should be a safe distance would appear to be inapplicable. A plate glass window in a shanty structure on the bin floor shattered and the fragments caused injuries. Within elevators only shatterproof, transparent materials should be used. In a large elevator where there are many employees there must be an explosion-resistant emergency warning system to call for an evacuation if there is time. In addition, provisions must be made for locating personnel after a disaster. An assembly area should be designated in order to identify the missing. Time cards can be used for employees but gate records must be used for farmers, truckers, salespersons, etc. Some idea should exist as to the previous locations of missing personnel. Power, water hydrant, and telephone lines to the elevator should be blast resistant. This probably implies burial at some distance from the elevator foundation. Employees should wear fire-resistant or fire-retardant clothing over all portions of the body, including the head and hands, to minimize the severity of burns. This is perhaps most easily done by using coveralls or Jump suits in conjunction with caps and gloves. Under no condition should employees go shiftless, wear only undershirts' or wear synthetic fiber (edge, polyesters) clothing, which considerably exacerbates burn injuries. Complete cooperation by management is most helpful in conducting an investigation. They can furnish blueprints, witness statements, flashlight batteries, and hospitality. One witness who was badly injured had been in a location where he could have clearly seen the course of the explosion but declined to cooperate on the advice of legal counsel. He was concerned that he would jeopardize any chance of financial compensation for his injuries. It is not known whether this difficulty is real or perceived-, but whatever the reason it can be a significant impediment to investigations. Two visits were made to the site of the explosion. One was 2 days after the explosion and the other after 2 months. Both resulted in valuable information. A third meeting with members of the elevator management was made some 6 months after the explosion. Because of the size and complexity of the explosion it was not until after the third meeting that a reasonable scenario could be assembled. Incident No. 2 An explosion occurred at a large wooden country elevator. There were rail loading and truck dump and truck loading facilities. There was a separate headhouse with two metal-encased bucket elevators. Two storage annexes were located on either nice of the headhouse and two metal grain storage tanks were at the end of one annex. The level-of housekeeping, as Judged qualitatively by the subpanel, was not good.

1 22 Observations The explosion occurred at mid-afternoon. Moderate damage was done to. the metal-clad, frame-structure elevator buildings. The one annex building seemed to suffer slightly more damage than the other. The bins were essentially empty. An explosion propagated down the tunnel destroying the door.at the end. Some of the bin bottoms were blown inward. The explosion was able to enter the empty bins in this fashion. The metal tank beyond the door was slightly burned on it8 side. The gallery and bin roof on this annex building were mostly destroyed. It had been displaced to the side and some of it was on the ground. The wooden bins were still intact but showed 'evidence of internal burning. The bin roof and gallery of the other annex were still intact but the gallery showed,some bulging. This annex was mostly full and the explosion did not enter the bins. The two sides of the headhouse that did not face the annexes were blown off from the bin floor level to the roof. There was little damage to the bucket elevators and related equipment. A slight amount of burned grain was present. Before the 'explosion, the elevator had been unloading trucks of soybeans. Someone smelled smoke coming from the boot pit area that was covered by a hatch on the elevator work floor. The fire department was called and the elevator was evacuated. Upon the arrival of the fire department, three elevator employees and three firemen lifted the boot pit cover and observed flames. The fireman with a baseline set the nozzle on fog and started to spray water into the burning boot pit. A "woosh" was heard and a ball of flame shot out of the pit'. The people who were around this area promptly exited the elevator and after they had cleared the facility a blast was heard coming from the elevator. The fireman began to put water on the resulting fires, which were mainly confined to the grain. The watering was stopped when it was decided that the burning grain could ' tee shoveled from the elevator. There was little fire damage to the structure. Of the six persons standing around the boot pit only one was injured during the attempt to escape. His exit was by an indirect path to the outside that took him through part of the elevator, rather than by the direct exit door to the outside that the others used. This slight time difference was apparently enough to allow the,expanding fireball to catch up with him. Scenario The boot pit area beneath the work floor had not contained a light ~ fixture and a regulatory agency had ordered the installation of one. This area had not been recently cleaned (probably for a month) and the accumulated dust had buried the light fixture and the light was turned on. The hot fixture, well insulated by the grain dust, caused the dust to ignite. The application of the fog water stream dispersed the burning and unburned dust into the air and an explosion occurred. This partially vented into the work floor area but also traveled up the two bucket elevators to the top of the headhouse, into the two galleries, and down the one tunnel.

23 Conclusions . Proper and correctly installed electrical equipment can cause problems if the surrounding environmental conditions change. The accumulation of dust not only presents a fire hazard but also an explosion hazard. In elevator designs, all areas, boot pits included, must have easy access, be well lighted, and be conducive to the performance of routine tasks such as housekeeping. Dust must not be allowed to accumulate in hidden areas where its danger goes undetected. - The fighting of elevator fires is a sensitive procedure. There seems to be a general belief that a fog stream will not disturb the dust and lead to an explosion. Until better methods are devised it is probably best to remove burning grain and dust from the elevator by the use of buckets ant shovels. Fire department personnel in areas where grain elevators are located must be educated concerning tactics for combating fires in grain elevators. Wooden elevators would appear to be leaky enough that the overpressure resulting from the explosion cannot build up to a high enough level to do major structural damage. Therefore, if the few resulting fires from the explosion can be initially controlled, the wooden structure need not burn down. The response time of the fire department is critical. When bins are empty the most severe explosion damage may occur. Every effort should be made to ensure that the bins are substantially sealed off from the tunnel, that the bin bottoms cannot collapse upward, and that the bins are sealed off from the gallery or distributor. Although the employees were aware of the possibility of explosions, none of them seemed aware that a dust explosion could literally tear an elevator apart. It is important that the grain industry trade organizations, the agricultural extension services, and the government regulatory agencies publicize this-problem. ~ - Incident No. 3 A fire and explosion occurred at a small country elevator and mill combined in the same wooden structure. There were rail loading and truck dump and truck loading facilities, the latter of the drive-through type. A wooden flat storage shed and an office building were located near the elevator mill building. A metal bucket elevator with twin legs ran through the central part of the structure. There was no headhouse structure as the head pulley was located on the roof. The bins had open tops. The level of housekeeping, as judged qualitatively by the subpanel, was not good. ! Observations The explosion occurred late at night. Minor structural damage was done to the wooden structure by the blast wave in the mill area, which was not well vented. A slight separation of the roof and walls occurred. Substantial fire damage occurred in the cupola where there was a sustained fire before and after the explosion. Slight charring of the wooden

1 24 structure occurred in the truck dump area and in the mill. Substantial structural damage due to the blast occurred to the 'lea caning in approximately the lower one-half portion. The belt had parted and dropped. It had burned through, but did not look as if it had caught on fire from frictional heating. Another area on the belt also showed evidence of burning. Prior to the explosion, milo had been ground and screened, mechanical repair n had been made on the screener, and wheat had been loaded into a truck. This activity had gone on from mid-afternoon until late evening, during which time the lights in the elevator had been left on continuously. At the completion of the loading, the 2 workers closed the elevator and went to the office to relax before going home. As they prepared to leave the office they noticed a flickering light in the cupola window and identified it as a fire. The workers called the fire department and proceeded to the roof (approximately'30 feet off the ground) on the external ladder to fight the fire with small hand fire extinguishers. When they looked through the window of the cupola they saw a fire raging on the bin floor ant on one wall where there had been a lighting fixture. They discharged the fire extinguishers with ' little effect and the one worker left to get more. An explosion occurred. At approximately this time, the fire- department arrived and successfully suppressed the continuing fire in the cupola. A fireman who was responding to the alarm was approximately 100 yards away from the elevator at the time of the explosion. He heard a "woosh" sound and saw a fireball come out of the cupola', out of the one open truck dump door, and out from around a large banging door on the feed mill. The resulting fire was successfully suppressed. Scenario A vapor-proof light fixture was installed on the cupola wall with its axis of symmetry perpendicular to the wall. There was as mucks Hi; `-- ! inches of dust-on horizontal surfaces including the top of the 'sit_ f ixture . The light f ixture became hot enough to ignite layered dust and this in turn fell onto the dust layer on the floor and ignited it. The wooden bin floor surrounded the metal leg casing. The fire surrounding this casing caused the bucket elevator belt to catch fire and burn through. Since the one side of the leg was closer to the ignition source than the other 8 ice, the latter had only started to burn by the time the belt dropped. When the belt with its metal buckets fell down the legs accumulated dust was disturbed, creating a combustible dust-air mixture in the presence of burning belting or sparking metal cups. The leg casing exploded, discharging a fireball into the truck dump area. This fireball vented out the one open door and into the feed mill where additional layered dust was stirred up, producing a secondary explosion, which vented through the hanging door on the feed mill. The fire in the cupola still continued out of control. Conclusions An accumulation of dust on the vapor-proof light fixture probably led to this accident. Presumably problems had not occurred before because the

25 lights had not normally been left on for this length of time. Had there been a sprinkler system in the wooden elevator, the fire may have been initially contained. Had not the leg belt dropped, an explosion probably would not have occurred. In this case an outside ladder allowed the workers access for fire fighting. Perhaps they would have been more successful in fighting the fire had a standpipe been available. Considerable care must be exercised in the fighting of dust fires because it is easily possible for the fire to become an explosion. The firemen were successful in containing the fire because they were equipped with a snorkel truck and the elevator was not very tall, approximately 30 feet. The blast damage to the structure was not severe because of the large venting area afforded by the one open truck dump door and the hanging door. Also, wooden structures are for the most part rather porous. Because of its low density, wood does not make effective projectiles. - The subpanel investigators arrived at this elevator a week after the explosion had occurred. It had not been reported to federal authorities. The accident was noted in news accounts. It is necessary that a system be instituted so that the occurrence of an accident is promptly reported. In some investigations equipment other than that normally carried by the investigators is required--ladders, boots for deep water, and winches. (In this particular case the fire department provided such equipment.) It is desirable not to let witnesses confer before they are interviewed. It is also best to interview witnesses first separately and then as a group. The investigators should be able to spend as much time at the site as required for a thorough analysis. Some investigation tasks, such as the removal of a dropped belt, are physically impossible for investigators to attempt. Therefore, it is important that an investigating team have the means to obtain the necessary assistance to handle heavy manual tasks. Incident No. 4 An explosion occurred at a country elevator of medium size with an associated mill. The elevator was of concrete construction. It had truck loading and unloading facilities and rail loading facilities. There was a single set of silos with a large Butler bin at one end and the headhouse and mill building at the other. There was no gallery. There was a screw conveyor in the tunnel. Housekeeping, as judged qualitatively by the subpanel, seemed adequate. Ob servations The explosion occurred during morning operations. Severe structural damage was limited to the concrete structure under the headhouse. This structure was external to the grain silos and contained the manlift shaft and the legwells. Approximately the top one third of two of the walls had been totally blown away. One of the damaged walls was opposite the bin wall and the other orthogonal to it. Moderate structural damage was sustained by the headhouse. The headhouse was connected to the leg wells and manlift

- 26 shaft through hales in the floor. The walls of the headhouse were bulged and the casement windows in the headhouse were broken and blown open. The boot area was full of rubble resulting from the collapse of the upper leg- wells and manlift shaft. The tunnel area showed evidence of burning and the door at the far end of the tunnel was destroyed by the overpressure. At the time of the explosion a rail car was being loaded with corn, and alfalfa pellets were being unloaded at the feed mill. The corn was being screened and trouble had been reported with the screener clogging. Subsequently it was reported that the entire system had plugged--screw conveyor, bucket elevator, and feed spout. Several witnesses heard the explosion and said that there was only one. Only one individual actually saw the explosion and he stated that the north side of the structure under the headhouse blew out spewing concrete and two bodies. The witnesses stated that 3 days before the explosion a fatal accident had occurred causing substantial mechanical damage to the bucket elevator. During the day preceding the explosion considerable time was spent repairing that damage. Observations made by OSHA employees subsequent to the team's first visit and during elevator reconstruction revealed that the bucket elevator belt was covered by concrete rubble in the boot area. It had metal buckets and was parted. At the break in the belt there was evidence of combustion. The lagging in the head pulley was only partially intact showing signs of being burned off. Scenario The bucket elevator was jammed either because of the mechanical damage that had been previously done to it or because of a choke brought about by the clogging of the screener. An attempt was made to "jog" the leg. It is not clear who turned the motor on as the only controls were at ground level outside the elevator near the boot. The two injured employees were blown out of the headhouse and the employee in the feed mill stated that he had not been in the elevator for the 5 minutes preceding the explosion. The slipping belt on the head pulley caused a friction fire that burned through the belt. The belt then fell down the legwells and dislodged accumulated dust. The flaming belt end or the metal cups striking the concrete wall acted as an ignition source for the suspended dust-air mixture. An explosion then occurred in the legwell and propagated upward into the headhouse. -The headhouse then exploded, venting through the headhouse windows to some extent. Also, the leg explosion was partially reflected off the headhouse floor causing the top portion of the legwells and manlift shaft to fall. This vented the high pressure gases. The explosion also traveled down the legwell into the boot. The explosion vented into the tunnel, in which it appeared an explosion was not sustained. Conclusions - This accident may have occurred as a result of events 3 days before. The elevator equipment had been damaged and the employees were still thinking about the fatal accident and the funeral. Generally, accidents occur more frequently when employees are under stress or distracted--Monday mornings, Friday afternoons, shift changes, graveyard shifts, overtime, etc.

27 If modern devices are employed it should be possible to prevent a leg choke. The leg feed should be controlled by a torque-sensing device on the drive motor. As the torque required increased, the feed rate would be decreased. The jogging of legs should be physically prevented through the use of interlock devices. The elevator boot should be easily accessible so that if a choke does occur it can be cleaned out. Shovel size access ports should be put in the boot, and the boot pit should be well lighted and free of obstructions. Provisions must be made to remove the excess grain from the pit. Concrete legwells allow little explosion venting. Pressures up to one atmosphere can probably develop before failure occurs allowing continued propagation of the explosion. When they do fail, concrete fragments act as projectiles and high pressure gas is released. In this case, however, since the legwell was on the exterior of the structure, the damage occurred to the surrounding area and not in the interior of the elevator. The truck dock was unfortunately within the distance of travel of some of the fragments. The destruction of the manlift shaft broke the only connection between the top and ground level of the elevator. The only access to the roof was by a crane bucket. Secure, blast-resistant means of egress must be provided so that injured employees can be removed from the top of an elevator. The bin bottoms and gates were of particularly substantial construction. Although a combustion wave did travel down the tunnel, it did not get into the silos and cause them to explode. Also, the absence of a gallery to some extent prevented the explosion from entering the bins. It is possible, however, for a distributor to propagate the explosion into the bins through the spouts. The arrival of the subpanel members within 24 hours of the explosion aided considerably with the investigation. Some public safety officials--State Fire Marshall--were still available. Furthermore, the witnesses were still willing to talk freely and events were still quite fresh in their minds. An investigator was able to return to the site during the period of reconstruction. He was able to examine items of interest as they were removed from the rubble. At the time of the first visit the bucket elevator belt and the head pulley were not accessible. During return visits it was also possible to visit with injured personnel who were not initially available because of hospitalization. Incident No. 5 An explosion occurred at a large concrete country elevator. There were truck dump and truck loading and rail loading facilities. A very large headhouse was located on the top of the silos and approximately in the center. There were two concrete double-leg bucket elevators, one used for unloading and one used for outloading. The truck dump was of the drive-through type located toward one end of the structure. An abandoned brick soy mill was at this end of the elevator. At the other end of the ~

28 elevator was a large steel storage tank. Along the one side of the elevator was a wooden building used for storage. The office was in a separate building some distance from the elevator. The level of housekeeping, as judged qualitatively by the subpanel, was poor. Observations The explosion occurred in early afternoon. Slight damage was done to the ad jacent wooden storage building whose end was parallel to the drive entering the truck dump. The blast, which came out of the truck dump, removed some of the weatherboard structure from the framing. The soy mill also received slight blast and missile damage resulting from high pressure gases venting from a personnel door on the end of the elevator. Fire, blast, and missiles moderately damaged the hopper-bottom trailer that was . . . standing on the truck dump. The fire damage seemed to engulf the truck; however, the blast and missile damage was the worst on the side toward the bucket elevators. The truck dump area showed a uniform scorching due to fire. Much grain from broken house bins had fallen into this area. The elevator control room adjacent to the truck dock and the leg was also burned by the passage-of a fireball. Both bucket elevators were severely damaged by blast in the vicinity of the truck dump. The legwells for each elevator were located on either side of the manlift-ladder shaft and the up- and down-legs were separate. Both down-legs had shattered completely, spewing concrete fragments into the control room and across the truck dump. The leg casings showed decreasing damage to the level of the screener f lo or approximately half way to the top of the elevator. The failure of the leg casings made both the manlift and ladder nonoperative. Fumigant had been stored near the legwells and as a result of the explosion had been dispersed around the work floor and truck dump area. The tunnel under the bins toward the steel storage tank showed slight evidence of an explosion. At the far end some damage had been done to the auger for the steel storage tank. Moderate damage was done to the tunnel area going to the other side. The tunnel terminated in a room with a personnel door and this door had blown off. It was clear that a rather large explosion had occurred in the truck dump area but to a large extent was able to vent through the open entrance and exit doors. Access to the upper part of the elevator was difficult as -the ladder in the manlift shaft had been destroyed. Entrance was gained by climbing to the top of a-grain drier that was located on the side of the - elevator opposite that of the storage shed. At this level, which was approximately 50 ft above the ground, there was also a window in the elevator. A 20 It extension ladder was laid from the top of the grain drier to the window. This was the level of the screener. Looking into the house bins it could be seen that the legwells had also exploded into the bins. The damage to the legwells was below this levele From this level it was possible to use the ladder to the headhouse . At the bin f loor level there was a Mayo spout coming from the distributor with a circle of holes for the respective bins. The floor was covered with much accumulated grain and dust. The walls were covered with dust to such a thickness that it almost resembled a layer of cork. A ladder and catwalk encompassed the distributor f loor and at the highest level within the headhouse was the head pulley and

29 drive motor. The head pulley covers had been blown off. The lagging on the unloading leg had bunched on one side of the head pulley. This caused the belt to slide to one side and it and the metal buckets were rubbing against the leg casing, which was metal at this point. The headhouse was moderately damaged. The broken windows and bulged sides indicated that an explosion had occurred within the headhouse. An inspection of the roof showed that the headhouse had apparently separated from the roof and that it had slid along spouts passing diagonally through the walls. It thus appeared as if the headhouse had to some extent lifted vertically. The roof on the end of the elevator toward the steel tank appeared to have hinged upward to relieve the pressure when some of the tanks exploded. Witnesses reported that it - neemed as if the entire roof of the elevator had lif ted . All of the bin covers were displaced. The spout feeding the large steel tank had fallen. Many of the spouts within the elevator were worn through and were patched with rags, boards, and wire. Fires had occurred in many of the tanks. At the time of the explosion rail cars were being loaded with corn and corn was being unloaded from the hopper bottom truck. The truck driver was apparently closing his hopper bottom from the side of the truck toward the bucket elevators. An elevator employee was in the control room, and another at the scales whose controls were located in the wooden storage shed. Scenario The metal cups rubbing the metal leg casing near the head pulley caused either a hot spot or sparks and this leg exploded. This explosion then propagated down the leg and into the distributor. After accelerating down the leg, it blew out at the work floor and boot pit level causing additional explosions in the truck dump area, the dump pit, and the two tunnels. From the distributor it was able to get into the second leg, the headhouse, and the various silos. Conclusions Out-of-alignment belts in bucket elevators may act as ignition sources either because they rub on a metal casing or their metal buckets strike a metal casing. Belt alignment monitors should be used and plastic buckets should be considered. Regular inspections should be made of the head pulley to check the conditions of the lagging. Inspection should be facilitated through the use of conveniently located doors, catwalk, and ladders. Through the application of suction it may be possible to remove the explosive dust concentration within the leg so that if an ignition source toes appear, an explosion need not result. Headhouses should be designed so that all horizontal and vertical surfaces may be regularly swept to remove dust. Elevators should be designed so as not to have interior concrete legwells. These allow explosion pressures to rise to a high level, vent high pressure gases into confined spaces such as bins where additional explosions occur and spew out concrete fragments. All too frequently these i

: 30 are slip-formed to include the elevator or manIift and ladder. These are, of course, also destroyed when the casing fails. It is necessary to provide ladders giving access to the top of the elevator in other locations. Control rooms should not be located inside elevators especially next to legwella. As far as possible personnel should not be in the truck dump area. With hydraulic lifts this is possible to some extent. The elevator roof should not be tied to the tank structure. It would then be possible for the high pressure gases to lift the roof and not cause a failure in the side of the tank. Open truck dump doors are very effective venting areas for explosions. . . . Although some concern has been expressed about ignition by hot surfaces of a running or recently run truck, it seems unlikely that a significant hazard exists. A surface must exceed 400°C and be exposed to a cloud of dust with' a concentration above 'the lower explosive limit. It seems unlikely that this will occur when a truck dumps or is dumped because the heaviest clouds are at the grating and the truck body shields the exhaust system from the 'high concentration dust clouds. , ' Incident No. 6 An explosion occurred at a large concrete country elevator. There were truck loading a,nd'unloading facilities as well as rail loading facilities. There was a large flat storage shed near the elevator and the office building and scales were separate from the elevator. The elevator was of unusual design in that there was no headhouse or gallery and the silos were arranged in a single circle about a steel-cased double-leg bucket elevator. Another elevator was external to the bin structure. Additionally, the entire elevator interior was slightly pressurized to help prev,=~ tne --e of dust into the elevator. The level of housekeeping, as judged qualitatively by the subpanel, appeared to be adequate. ' Observations The explosion occurred in mid-morning. There was a fatality and several injuries. Severe damage occurred to the elevator control room, the' penthouse that covered the interior legwell, and the truck that was located- on the truck dump. Moderate damage occurred to all leg casings and the flat storage shed. A significant explosion occurred in the basement of the elevator and the escape of these ' high pressure gases caused the severe damage. Part of the gases went up the center of the circle of' tanks where the outloading bucket elevator and the manlift were located. This destroyed the penthouse, which was situated on top of the vertical shafte Six-inch steel channels that supported the structure were bent by this blast. Additional high-pressure gas escaped through the tunnel to the truck dump. The elevator control room was located on top of this tunnel between the elevator structure and the truck dump. The floor of the control room was lif ted to almost ceiling height when the top of the tunnel failed. Burned gases e scaped around the metal dump pit and through the hydraulic truck

- 31 hoist and lif ted the truck that had just dumped . The remaining high- pressure gases vented through a personnel door forming a high-velocity Jet that impacted upon the flat storage shed approximately 75 feet away. Some of the 2" x 4" structural members were broken and some of the sheet metal covering wan torn off. The damage to the interior outloading leg was caused by the belt dropping. The ends of the belt were badly burned and the head pulley lagging was partially missing and showed evidence of combustion. The outside unloading leg exploded causing the metal casing panels to separate at the corners. Some of the bin covers had been lifted of f and there had been fires in several of the bins. Just before the accident, corn was being loaded into a string of rail cars and corn had been dumped into the truck pit, but it was yet to be elevated. The interior leg, which was feeding the rail cars at this point, choked. Help was sought to dig out the choke, but none could be found. A decision was made to Jog the leg and thus defeat the interlocks. A drop in amperage of the bucket elevator motor caused the supervisor to go outside and look at the head pulley and then go to the boot area and open the inspection door to see if the belt had dropped. The belt had not dropped and the supervisor took the manlift to the top of the elevator to inspect the head pulley. At this time the explosion occurred. The supervisor was approximately two thirds of the way to the top when struck by the blast. He managed to get out of the manlift and up the inside ladder to the top of the elevator and then proceeded to climb down the outside ladder. A farmer in his truck cab was slightly injured when his truck was displaced by the blast. The elevator employee at the truck tail gate was thrown into an adjoining field by the blast. The elevator operator in the control room was crushed when the control room floor was lifted to the ceiling. The local fire company arrived at the scene, strung hose lines to the 100 foot high elevator roof, and extinguished the bin fires by hosing down the grain. Scenario For some reason a choke occurred in the outloading leg. As help was not readily available to dig out the choke, it was decided to jog the leg . This caused a belt and lagging fire at the head pulley. The boot inspection door was opened, allowing pressurized air from the elevator interior to flow into the leg and escape to the outside. This flow of air intensified the belt and lagging fire. The belt burned through and dropped down the leg. For some reason the leg did not explode. The collapsing belt pushed a cloud of dust out of the open boot inspection door and caused the leg casing deco split. I Burning pieces of lagging and belt fell out of the casing and into the dust cloud in the basement of the elevator, which subsequently exploded. Cone fusions The bucket elevator choked and in spite of interlock devices an attempt was made to jog the leg. Chokes can be prevented with motor torque or belt tension devices controlling the bucket elevator feed gate. The boot area

32 was quite accessible, so it should have been relatively easy to have removed the choke. Employees must be educated that jogging a leg may be a fatal mistake . The blast demolished the control room and the employee lunch and break room. Areas where personnel normally congregate must be removed from the elevator. Since escaping jets of gas from doorways and windows can travel large distances, buildings should not be placed opposite the elevator. It is clear that fighting fires on the top of an elevator is difficult and may not be desirable from the point of view of grain salvage. Dry standpipes to the top of the elevator are probably desirable, but for a severely burning bin fire it is not clear that water is effective. Indeed combustion of grain in a limited oxygen environment may produce carbon which then reacts with water through the water gas mechanism to produce quite dangerous carbon monoxide and hydrogen. It still remains to be determined whether water will effectively penetrate a bin fire. The water-logget grain also expands and if not quickly removed can place nonde sign mechanical loads that lead to failure on the silos. Bin covers may, if large enough, effectively vent the high-pressure gases from exploding partially filled tanks. Silos that are nearly full do not explode effectively. However, the bin covers must be securely fastened to the bin tops with a short piece of cable or chain or they become effective missiles. Distributors are effective in multiplying the paths of explosion propagation as numerous spouts are interconnected at this point. If an explosion comes down from the bucket elevator feed, it can usually get into each bin. The truck dump is an effective large high-pressure vent for the basement areas. (Personnel located in this area are likely to be injured in an explosion. Therefore, only personnel actually participating in the operations should be in this area. Truck drivers and visitors should be encouraged to wait elsewhere.) During this investigation, cooperation with personnel from the investigative agencies was good. Also, additional people had been brought in to assist with elevator operation and repair; thus, the work load was considerably lightened and employees were able to spend much more time with the subpanel investigators. Blueprints and flow diagrams provided to the investigators aided the investigation considerably. i\ 1

Guidelines for the Investigation of Grain Dust Explosions Get This Book
×
 Guidelines for the Investigation of Grain Dust Explosions
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!