Click for next page ( 14


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 13
Chapter 3 THE EXPLOSION PROBLEM Grain dust suspended in air at a concentration above a certain lower limit* burns rapidly when ignited. When thin occurs in an enclosed space, explosive pressures are generated and detonation may even occur. The panel found that the potential for grain dust explosions existed in every elevator and mill it visited. An extrapolation to the approximately 15,000** elevators existing in this country at present indicates the potential magnitude of the problem. Data on recorded explosions are presented in Table 1.*** Any movement or handling of grain produces grain dust, and this occurs at each point in the grain-handling system from farm to ultimate consumer. In the past 20 to 2S years the grain-handling industry has processed a continuously increasing amount of grain from U.S. farms. Between 1964 and 1979, for example, grain production increased from 160.7 to 280.7 million metric tons and the amount of grain exported (U.S. Department of Agriculture 1979a) exhibited an even greater increase from 41 to 113'million metric tons. Corn accounted for the greatest portion of both of these increases. Although it is difficult to obtain exact figures for the amount entering the grain-handling system (i.e., total production minus on-farm use), there is no doubt that it has increased approximately 150 to 200 percent. This increase indicates that the velocity of flow through the grain-handling system has increased, that the units of the grain-handling system have increased in size, or that both have occurred. These changes have increased the explosion hazard since the amount of dust generated in a facility increases with the total amount of grain handled and with increased operating equipment speed. * The lower explosive limit (LEL) range according to U.S. Bureau of Mines 1961 data for some common agricultural dusts is from 35 to 300 g/m3 (0.035 to 0.29 oz/ft31(See Table B-1 of Appendix B). See Appendix B for a detailed discussion of the explosive properties of grain dust. ** Includes all installations involving the elevation of grain (i.e., conventional elevators and elevators attached to mills). *** In addition to explosions, over 29,000 grain elevator fires occurred in the U.S. from 1964 through 1973 or an average of more than 2900 fires per year. This is reported to be more than 300 times greater than the average number of grain elevator dust explosions in the same period (Verkade and Chiotti 19761. 13

OCR for page 13
14 .. . . . . t O 0 . ~ ~ ~ ~ 0 ~ . Am. O ~ ' ~ ~ ~ ~ 0 . ~ _' ~ D ~ _. O , . ~ _I IN ~ O _I .a ~ 0 , ~ ~ ~ ~ X ~ ~ V - - ' ' 0 ~ ~ O ,,, . ~ ~ 3 , ~ U-) O A ~ ~ ~ O . - ~ ~ ~ t) 0 ~ ' ' ' V _' O at: art ' O ~ ~ 0 ~ . . ~ ~ S . - . 0 ~ ~ ~ . ~ ~ O O ~ 1 1 1 1 1 1 1 '1 1 1 1 un _' ~ ~ ~ ~ ~ 0 ~ o ~ us O ~ ~ C: if. _ ~ ,4 ~ ~ ~ Cat ~ ~ ~ ~ ~ 0 ~ 'A O 0 ~ ~ ~ ~ ~ O ~ ~ ~ ~ a' . , ~ ~ 0 . c, ~ ~ C) . X ~ O ~.0 . . ~ ~ O 3 ~ - o ~ ~ ~ ~ ~ 1 ' ~ 0 10 ~ ~ O . .. . . U ~ ~Q . . ,~ ~ ~ ~ 0 ~ ' S- 0 P4 ~ ~ 0 cn ~ ~ O m ~ - - t) 0 Ll 0 0 ~ O ~ ~ a' ~ 0 i` ~ ~ ~ ~ O . O ~ ~ U) ~ ~ ,_l o~ tU ~ ~ _ ~ O ~ ~ ~ O . ~ ~ ~ ~ ~a . ~ `: ~ ~ I . . a, ,~ ~ .. V X _. , _ ~ ~ :> cr' ~ ~ ~ ~ ~ ~ ~ ~ ~ un ~ u, o ~ - .~ u~ ~ ~ a, ~ ~ o _~ t~ 1 1 ~ LO O _t _' ~ . dJ ~ ~ ~ `: 0 ~ ~ U' ~ ~ . ~ .,' ~ ~ . _I ~ ~ ~ ~ ~ X O ~ 0 ~ ~ ~ - . .- ~ ~ 0 ~ ~= 05) 3 3 ~: ~ _ . ~ U~ ~ ~ ~ O 1- ~ ~ ~ 'O : . _ . ~n ~ ~ ~ ~ ~ 40 ~ ~ ~ ~ 0 0 ~ 0 e ~ ~ ~ ~ :> 0 u 0 _ 0 ~ ~ ~ ~ = O ~) ~ 3 ~ ~ aQ Q~ _~ , P. ~ U ~ O ~ ~J m ~ - ~ u P. en ~ O ~ ~ ~ CO C~ ~ ~ ~ ~ 0\ ~ ~ ~ O ~ C~ ~ O ~ CD~ r~ ~ ~ O C _~ _~ _1 _~ _I ~ _~ I I c a ~ c5 _I O ~ ~ ~ 0 X ~ . ~ c:' ~ ~ ~ u ~ ~ a 0 O U) O " - ,.' ~ ~ ~ ~ r~ 3 ~ O _ ' ~ C) O ~ X ~ . ~ o~ 0 0 0 - 0 0 ~ ' S V C X ~ ^ - 0 C~ ' ~ ~ 0 ~ (U O ~ ~ a' ~ ~ ~ cx, cn ~ ~ o ~ o ~ ~ o ~ ~ ~ tD ~ O U~ 1i: ~ ' _~ _~ _~ _~ _~ ' _t _~ 1 1 1 c: ~ o ~Q ~ O ~ _I ~ cn ~ 3 Q, ~ . ~ .- ~ ~ V ~ O ~ tD ~ 0 ~ cn =~ ~ ~ = - ~ ~ .O . ~ ,~ ~ ~ ~n Z ~ _ , S0 ~ ~ ~ ~C E~ ~ ~ ~ O O _I C tQ >, . e ~ AS ~ ~ . ~ er O _t ~ ~ ~ U~ tD ~a~ ~ o _t c~ ~ ~ ~ ~ ~ co ~ o ~ o O ~ ~ ~ O U~ ~ (D tD tD tD,~D tD ~ ~ tD ~ ~ ~ ~ ~ ~ a~ ~ ~ P~ ~ ~ ~ ~ ~ . ~ 0 o~ ~ ~ ~ ~ ~ ~ ~ ~ cr~ ~ c~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0 ~ 0 a n~ ~ ~ C E~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Z ~ ~ ~ u: ~ ~ ~

OCR for page 13
15 Other innovations or changes also have tended to increase the amount of dust in the system. The rapid drying of grain on the farm or at elevators by the application of external heat leads to differential shrinkage, which increases the tendency for the kernel to break. The increased foreign demand for corn and soybeans has changed the mix of grains passing through the system, and although there is no definite proof, it is thought that these grains are considerably dustier than wheat, which was the main export grain prior to 1960. Minimizing the handling of grain will reduce the overall hazard because dust is generated every time grain is moved. The movement of grain through the handling system is much less seasonal than in prior years due to the large increase in on-farm storage facilities. Available data do not support a correlation between the incidence of explosions and any of these factors. In addition, extrapolation to predict any future hazard due to external causes is next to impossible since the characteristics of each seasonal year (i.e., weather, crop quality, export market, etc.) are unique. The panel, as well as others who have studied the problem as far back as 1918, is of the opinion that the underlying and by far the most important hazard is grain dust itself. In an operating elevator or mill, grain dust is Generated and handled in confined spaces, and without proper housekeeping it will accumulate in layers on interior as well as exterior surfaces. Although there are a number of other factors (such as presence of ignition sources), which contribute to the degree of hazard, the panel feels that they are secondary compared to the hazard imposed by the accumulation of layered grain dust on interior surfaces. This is because the layered dust is the fuel for secondary explosions that occur. GRAINHANDLING FACILITIES Figure 3 illustrates the physical construction of an elevator (see Figure 2 for a flow diagram of the functions performed in a grain elevator). Not all elevators contain every feature shown but the illustration is a general representation of most of the activities that are performed. One must realize, however, that even with a minimum amount of handling between input and output there are a number of places where the grain is subject to mechanical stress leading to the production of grain dust. - Grain is delivered to an elevator by truck, rail, or barge and is dumped into a pit that feeds a conveyor belt leading to the bottom of the leg, which is called the boot. The leg is an enclosed, vertical, endless-belt, bucket conveyor that elevates the grain and discharges it into the top of a garner. Grain is discharged from the bottom of the garner into a scale bin in batch lots and is weighed. The grain then flows out the bottom of the scale bin onto a belt conveyor that moves the grain to the top of a bin or silo. The first grain dumped into an empty silo may drop as much as 100 feet or more. Grain is unloaded from the bottom of the silo onto a conveyor belt that feeds into the boot of the leg. After being elevated, the

OCR for page 13
16 . ~ I~~ 1 ~ ; ..~. > \ me - ~: . Z ~ ~ < o 1 1 .,~ 0 ~ ~ _ t ~ u o an' - V so ' . lo Cal o lo ~ ~~ ~ lo. i' o o . ~ ~ ~ _ a, a' U ~ a) ~~ ~ ~ , |, if ~ ~~ ~ o LIZ. To I i' - a ,- m 0} Hi a. ll, I'm i.= 3 ~ ~ Y "-i ~ ; = ~ u , o US a! V US Cd v ~ 3 i Cal bo ~ ~ o L \ o y o g At c) ~ ~ ~~ ^# ~ u .~:~r.~.' `' ~~' '

OCR for page 13
17 grain may take one of several different paths (e.g., it may go through the weighing process again and be loaded into a truck, rail car, or barge; it may be blended with grain from other silos and returned to a silo; or it may go to a drier). It should be noted that, except in rare instances, the grain is elevated in the leg at least twice during the time it is in the elevator and the leg is always operating when grain is being moved within the elevator. Elevators vary in size from 400,000 to 800,000 bushels for the average storage capacity of country elevators (some may be smaller or considerably larger) {T. E. Stivers Organization, Tnc. 1980) to an average of about 4 million for those terminal elevators registered under the Uniform Storage Agreement in 1978 (Development Planning and Research Associates, Inc. 1980~. Eighty-seven percent of export elevators have a storage capacity greater than 2 million bushels, with some exceeding a 10 million bushel capacity (U.S. Department of Agriculture 1980~. Newer elevators are of slip-form concrete construction whereas older elevators are usually of wood construction, the latter being mostly limited to country elevators. The geometry of elevators and the method of distributing the grain within the elevator also vary widely. Some elevators have legs external to the structure; some use inclined conveyor belts in place of legs, some distribute grain from a leg or inclined belt into a system of pipes (a distributor) leading to silos, thereby eliminating the gallery on top of the bins; some group the bins in a circle around the headhouse instead of placing them in a row as shown in Figure 3; some extend bins on each side of the headhouse; and some have combinations of two or more of these and other features even to the extent of having wood and concrete structures together in the same installation. In addition to the grain-moving system in an elevator, most newer and some older elevators have a pneumatic dust control system. Airborne dust is collected at various points in the elevator to improve the working environment, to reduce the hazard of dust explosions, and to meet Environmental Protection Agency (EPA) requirements concerning ambient air quality. In some elevators, the dust is collected in a separate bin and offered for sale; in others, the dust is returned to the grain. The efficiency, construction, and safety of dust control systems vary widely throughout the industry. Some mills for the processing of grain are attached to large elevators and milling is only a small part of the total elevator operation. The operation of those mills that are not adjuncts to grain elevators can be divided into two parts: the elevation and storage of grain and the processing of grain. The elevation and storage portion of the facility has the same problems with dust generation and control as any ordinary elevator. In both types of mill, grain processing (which can include grinding, milling, mixing, pelletizing, etc.) introduces additional explosion hazards. A discussion of the hazards peculiar to mills is given in Appendix C.

OCR for page 13
1 18 Proposals for reducing the explosion hazard must take into account the great variation in types and sizes of elevators and mills, in dust control systems, in types and amount of grain handled, and in the treatment the grain receives in the elevator or mill. DUST GENERATION Grain dust is generated initially during the harvesting process, and grain delivered directly to an elevator from harvesting contains an amount of dust that reflects harvesting conditions. Although all grain dust can be considered to present a hazard, the most hazardous in that which can become airborne. There is no particular standard defining the size of airborne dust particles; however, for purposes of comparing ignition and burning characteristics, the U.S. Bureau of Hines (1961)' used samples of various dusts, including agricultural dusts, that would pass a 200-mesh screen (maximum particle size 74 ~m). The degree of explosion hazard for a unit mass of dust is dependent on the total surface area of the dust, which, of course, is inversely proportional to the square of the size of the particles. me ease by which dust becomes airborne also is inversely proportional to the size of the particles for dusts of equal density. Dust suspended in air represents the initial hazard since dust clouds can explode'. A detailed discussion of dust cloud explosion phenomena is contained in Appendix B and an overview of explosions in grain-handling facilities is presented below. It should be noted, however, that the fuel for the dangerous secondary explosions is the layered dust that has settled on various surfaces in the elevator or mill. Dust on some surfaces such as floors, walls, beams, and ductwork is readily visible and can be removed easily. Neglected-'or delayed housekeeping will result in the accumulation of layered dust in an elevator or mill and poor housekeeping methods or an initial explosion can result in the creation of a dust cloud sufficient to be an explosion hazard. Unfortunately, housekeeping and maintenance often are given very low priority and usually are the first tasks postponed when there is a rush of business (e.g., elevators operating continuously during harvest season). Hidden layers on walls and floors of enclosed areas such as bins and ducts, however, have fueled serious explosions in apparently (supe'rficially} clean elevators. Horizontal surfaces inside of equipment should be eliminated through design. me points at which dust is generated and suspended in air can be identified by following the path of grain through an elevator as was done above. The panel has found that the dust concentration in operating elevator legs, even with currently designed dust control systems, is frequently above the lower explosive limit just above the boot on the up-side of the'leg. Dust concentrations ranging from 27 to 500 g/m3 were measured in this location in four elevators. Only slightly less hazardous concentrations were found to exist at other points in the leg. Panel investigators and others

OCR for page 13
19 have found that numerous explosions have resulted from the ignition of suspended dust in legs. Thus, the Panel considers elevator legs to be the most dangerous location with respect to initial dust explosions. The same - conclusion has been expressed in numerous previous reports (e.g., Wade et al. 19 80) . At the top of the leg or the inclined conveyor where the grain is discharged into a distributor or onto a conveyor belt, some of the dust entrained in the grain stream will be released when the grain falls through the air and additional dust will be generated when the grain impacts at the bottom of its fall either onto the upper conveyor belt or into an enclosure such as a bin or silo. Silos, bins, and garners are particularly hazardous locations since they are enclosed structures and, while being filled, contain high concentrations of airborne dust in a large volume. After being emptied, a layer of dust will usually cling to the walls. When grain is moved out of the elevator, it usually is discharged from a bin onto a moving belt at the bottom of the bin in a tunnel, another site for the generation of dust and the release of small particles that become airborne. Layered dust is particularly prevalent in tunnels below bins. The grain is conveyed to the boot of the leg and repeats the process of being elevated and distributed with the attendant opportunities for the generation of additional dust and the suspension of the finer particles. Other processes in the elevator such as drying, cleaning, and blending all require movement of grain through the leg and various amounts of conveying to and from the leg, again with the generation and suspension of dust. Thus, there are numerous places in an elevator where dust can be generated by mechanical stress on the grain kernels and where fine dust can become airborne and eventually settle as a layer on surfaces. Although the amount of dust that can become airborne at any one time is only a small fraction of a percent of the grain, over a period of time the total quantity accumulating on surfaces can be quite large.* A dust control system is the only method for preventing most of the airborne dust from settling on elevator floors, beams, duct work, etc. and clinging to such surfaces as walls, bin sides, and ceilings. DUST EXPLOSIONS* * Dynamics Discussions of dust explosions begin with the concept of an "explosion pentagon. whose sides represent the needed elements. The five sides are ignition, fuel (the dust), air (oxygen), mixing, and confinement. Although these are necessary requirements for a dust explosion, they are not sufficient. Explosions will not occur unless the dust is suspended in an enclosure in air at a concentration exceeding the lower explosive limit. These conditions must occur simultaneously and at one point--an enclosure containing dust suspended in air in the proper concentration with an ignition source. 1

OCR for page 13
20 Therefore, there are four events that must take place before a grain elevator will explode: . Grain dust must collect in the elevator. Grain dust must be suspended in air inside the elevator at a concentration above the lower explosive limit. Suspended grain dust must be ignited. Sufficient grain dust to sustain rapid combustion must be in the vicinity of the initially-ignited grain dust. The explosibility of dust under these conditions has been well documented (Bartknecht 1981, Palmer 1973, U.S. Bureau of Mines 1961~. These four events are shown in the fault tree analysis-of Figure 4. In this type of analysis, which is widely used to determine and estimate risks, all subordinate events leading to a major undesired event (in this case, Grain dust explosion in elevator.) are arranged to show their interrelationship and sequence (Rowe 1977~. These subordinate causative events may be independent or dependent. Independent events; e.g., Grain dust is blown down (by an air gun during housekeeping) by airstream,. lead into an OR gate, indicating that any one of such events can cause the next higher event. Dependent events; e.g., Grain dust is produced in elevator," require at least one other event to occur before the next higher event can happen. They are shown as leading into an AND gate. A peculiarity of explosions occurring in dusty structures is the phenomenon of secondary or subsequent explosions following a primary or initial explosion. me secondary explosion can be orders of magnitude greater and more destructive than the initial explosion. If the interior of a facility is dusty, an initial explosion that causes only relatively slight damage can produce a large, suspended dust cloud. Ignition of this cloud by hot particles of dust or flame from the initial explosion (or by any other ignition scurce) then causes the secondary explosion. Secondary explosions have been known to occur many seconds after an initial explosion. There also have been cases where a series of secondary explosions occurred, one after the other. * At a large export elevator during 11-1/2 months of operation 13,000 tons of dust had been removed from 1.8 X 108 bushels (equivalent to 4.5 x 106 tons assuming a bushel weighs 50 lb) of grain. ** See Appendix 8 for a more complete discussion of dust explosions.

OCR for page 13
21 m ree conditions for explosions--air, dust, and confinement--always exist in contemporary elevators and mills. The incidents described below illustrate how a sequence of events led to the simultaneous occurrence of all of the conditions required to produce an explosion: 1. During construction work welding wan being done on a conduit near an elevator leg that was built external to the headhouse. Some hot welding slag fell, unnoticed, into grain carried in an open screw auger to the leg. The hot slag was carried into the leg boot where it ignited the grain, which in turn, ignited the dust in the leg and caused an explosion. Because the leg was exterior to the elevator structure, no secondary explosions occurred. 2. During the operation of an elevator the bucket conveyor in the leg became jammed by an excess of grain in the boot. An employee began jogging the leg,* a procedure in which the leg conveyor driving motor is repeatedly started and stopped in an effort to free the conveyor belt. Heat generated by friction between the head driving pulley and the jammed belt ignited the belt, which subsequently burned through. When the burning belt dropped, the metal casing split. Flaming material was discharged into the dust cloud in the basement and there was an explosion in the basement and legwell shaft. Hot gases and burning particles were blown through a distributor system into various bins where they caused secondary explosions fueled by dust that was shaken loose from the walls by the initial explosion. 3. In an elevator like that shown in Figure 3, with silos extending out on each side of the headhouse, an initial explosion occurred in a load-out bin (also called a transfer or workhouse bin; see Figure 31. Explosions followed in the three legs in the headhouse and severely damaged the floors and interior partitions. The explosion vented out both sides of the headhouse at the top, across a catwalk, and into the galleries on each side. Dust that had settled on the floor of the galleries below the sides of the conveyor belts propagated a flame, with little pressure rise, down the length of both galleries. The flame travelled the full length of one gallery to an open silo at the end where a secondary explosion occurred. In the other gallery, the flame reached an interstice that was used as a ventilating shaft that extended from the gallery to the tunnel below the silos. An explosion occurred in this space when dust that had been shaken loose fran the walls by the explosion in the headhouse was ignited. This explosion then travelled down the tunnel with unusual violence, destroying the conveyor in the tunnel and the loading spouts, and vented out the far end of the tunnel. * Jogging the leg, although a practice prohibited in most companies, - probably occurs more often than industry is aware of or willing to admit.

OCR for page 13
1 22 o ._ - x 0 _, o 3 At C 0 ._ _ ~ C O 4" . ~ 3 3 .C C,) , tn 0 a' 3 c ~ . ~' 0 ._ 3 ~ c ~ r' 0 ~ /= _ D'~e ~] C .m . - 00 ~ . _ L c E =- ~ c~ -o 0 ~ 3 - ~ o ._ O O ~ 3 0 Q.= ~ 0 ~ =0 0 Q C ~ _ C .. m 0 ._ . s , Q~ O O O ~ -.C 0 ~ C ,,~ _ c~ o E o 0 c ._ ._ ~O r I ~ ... 8~= . ' ~ a, 0 C~ ~ O .C C, . - O 8 ~ ~ ~, C X ~ I 0 cn ~ ~ ////~ . o _ ~ ~ E E = e {,, {, - C - 0 C ~ E0 ~ ~ ~0 ~ 0 ~ 4- _ O ~ cn 3 ~n . ~ ~ 3 0 ~ ._ 3 Q ~ O o X ~ ~ L~ C ~ =2Q~ =0 ~ O ~ `~.- E ~ _ 0 ~ _ ~ 4,,, ._ ~ ~ O cn 1~ -e ,~ ._ _ ~o' (D Q. . . 0 o.C 'o ~ QD C . . O ~ Q D u7 {_ ._ - 3 ~ $ 0 ~ . o _ ._ . C~ 0 ~ ~ . D w O _ 0 ~ ~ m~ 0 _ o, ._ ~ ~ _ 3 ~ O .= 2 ~ C D 0 U, 1~. 1 1 ~ c) ~ 1 - --C 3 , ~ E ~ c _ ~ O s, ~0 Q _ 0 0 C~ 0 30 ' m ~ ~ ._ ~ c . _ ~ ~ 0 ~ _ _ ._ '', n 0 c~ ~ 0 1 . u, u, c~ s~ u c:

OCR for page 13
23 The latter two examples also illustrate the difference between initial and secondary explosions. Even though all of the conditions for explosion may exist at only one place at one time in an elevator or mill, an initial explosion can easily create these conditions at other points. The examples also identify the locations of the initial explosions. In same cases, secondary explosions have- been triggered by ignition of a dust cloud created within the conveying equipment by layered dust being disturbed and set in suspension in air. This hidden hazard of layered dust was illustrated by an explosion, investigated by the panel, that occurred in an elevator with ' remarkably clean working areas. In this case, 'the explosion was propagated mainly through hooded conveyors in the tunnels below the silos. Although the work areas had been kept clean by a combination of duct collection and manual housekeeping, quantities of dust sufficient to propagate an explosion had accumulated on interior surfaces of the covered conveyors. The explosion was violent enough to blow out the walls of the tunnel above ground and to initiate subsequent explosions wherever the explosion reached areas where layered dust had been shaken loose or scoured from inter) or surfaces by the pressure wave . The 'secondary explosion hazard due to layered dust cannot be overemphasized. LEL values from U.S. Department of Agriculture 1980 data for grain dust clouds range from 20 to 55 g/m3 . Tossing an LEL of 40 g,~in3 and an average density of 18.5 lb/ft3 for grain dust, a dust layer on the floor' of 1/64 in. or more in depth' in an enclosure 10 ft in height, can create a dust cloud above the LEL when uniformly dispersed throughout the enclosure. It also should be noted that an explosive concentration in a volume a few inches or feet above the floor (i.e., less than the full volume of the enclosure, can easily be created from a 1/64 in. layer by a slight breeze or other disturbance. Ignition of a small cloud such as this could disperse the remaining dust and result in an explosion throughout the volume. Ignition The initial explosion can be ignited by a multitude of sources. A recently published compilation of actual explosions lists 26 different ignition sources for 147 explosions and lists an additional 103 explosions as having unidentified sources (see Table B-2, Appendix B). Other publications show a s imilar multitude of sources and a large percentage of unidentified sources (for example, Bartknecht 1981, Palmer 19731. Although the panel believes that adequate investigation of the explosions attributed to unidentified ignition sources would have identified most of the sources (National Materials Advisory Board 1980), a valid ranking of the danger due to particular ignition sources based on present data of frequency of occurrence cannot be made because of the large number of unknown sources and because of the relatively large number of unreported explosions. Considering how easy it is to ignite grain dust, either through careless action or circumstances beyond reasonable human control, the potential for ignition exists in numerous locations and at many times in an elevator or mill. 1

OCR for page 13
1 24 The use of motors, junction electrical equipment not designed hazardous. Lack of (or improper) result in electrical sparks or in a cloud. The lack of ~ __ foreign objects into the grain stream can result in sparking, rubbing, or jamming, and the lack of means to indicate belt slippage and misalignment and switch boxes, lighting, and other for dusty locations can be extremely grounding of electrical equipment can sufficient to ignite grain dust, either layered a meabanism to Prevent the introduction of can result in friction fires or electrical arcing. Poor maintenance of equipment such as motors, bearings, and conveying systems easily can result in hot surfaces capable of igniting dust. Hot work such as welding or torch cutting and electrical work done during facility operation is particularly dangerous. Nonadherence to smoking regulations is an obvious danger. The effect of relative and absolute humidity and electrostatic phenomena on the ignition of dust clouds is a very controversial subject. Although the data relative to electrostatics are inconclusive, some believe that low relative humidity can contribute to the dust explosion hazard because the potential for electrostatic sparking increases as relative humidity drops and low relative humidity reduces the moisture content of grain dust, thereby lowering the ignition temperature and energy (Palmer 19731. As a consequence, employees of the Federal Grain Inspection Service (FGIS) now are advised to leave an elevator whenever the relative humidity falls below a certain percent and air-suspended dust is above a certain density (Federal Grain Inspection Service 19803. The panel believes, however, that the present basis for evacuation of an elevator needs re-examination. For example, a very serious explosion occurred at the Farmers Export Company elevator in Galveston, Texas, at 8:30 p.m. on December 27, 1977, when a relative humidity of 100 percent was recorded. The major difficulty with accepting data on past explosions is the fact that many of the values of humidity were obtained from readings taken anywhere from 5 to 100 miles distant from the elevator. Also, even assuming that local outdoor temperature and humidity readings could be obtained, the temperature in an elevator during winter may be high enough above the outdoor temperature to create a condition of very low relative humidity even though the outdoor relative humidity is high {e.g., air at 40F and 100 percent relative humidity, when heated to 70F, has a relative humidity of only 33 percent). During periods of low relative humidity there is a tendency for static charges to build up on nonconductive materials and this increases the possibility of electrostatic discharges. Admittedly, the potential for sparking exists in systems employing moving, poorly conducting belts, and electrostatic discharges are observed frequently in elevator legs. However, the panel questions whether the energy in electrostatic sparks generated in elevators or mills is released in a manner that ignites grain dust. The error of basing safety on high relative humidity is well expressed by the National Fire Protection Association (1978~: ....moisture cannot be considered an effective explosive preventive since ignition sources provide more than enough heat to vaporize and heat the moisture and ignite the dust.

OCR for page 13
25 There is a considerable body of documented laboratory evidence that electrostatic discharge can ignite dust clouds; however, in its investigations of explosions, the panel found no evidence of ignition due to an electrostatic discharge in an actual elevator. The panel therefore recommends continued research to elucidate better the role of electrostatic discharge and absolute humidity in grain dust explosions. CONTRTBIJTING FACTORS In considering accidents, including grain dust explosions, there is a tendency to focus on immediate physical aspect even thou gb accidents may be recognized as resulting from a combination of causative factors. Thus, the contribution of human operatives and external factors, singularly or collectively, often is overlooked. The panel'believes that the grain dust explosion problem in part results from people's attitudes and other seemingly unrelated factors. The attitudes of owners, operators, end ' employees that may contribute to the explosion hazard are discussed below. It must be noted, however, that the panel does not mean to imply that these attitudes are universal or that other equally dangerous attitudes do not exist; it only wishes to emphasize that the potential contribution of human attitudes must be recognized. Attitudes Owners. It may not always be in the best economic interest of a mill or elevator owner to protect his facility against explosion if it would require a capital outlay that is large with respect to the original cost of the facility. For example, an old facility may be insured for its replacement cost, which is several times its original cost; if it were to explode or burn (without injuring or killing anyone), it would be an advantageous method of replacement. Some owners of facilities with long explosion-free histories believe they have no need for concern. They may view government concern about explosions as an unwarranted intrusion into their business. Operators-Managers. The attitudes that elevator and mill operators may have about dust explosions are not unique and are held by some involved in other hazardous operations. Most prevalent is a slowly developed complacency. Even though a manager may intellectually acknowledge the danger of a dust explosion, heavy work schedules, emergencies due to equipment breakdown, worker absences, and other managerial pressures can mitigate his continuous sensitivity to the hazard. Even when operators hear of an elevator explosion elsewhere, they may believe that the accident was due to some extraneous or stupid action they would never commit. This is a difficult attitude to change without a method

OCR for page 13
26 for properly investigating and reporting on elevator explosions, and the panel believes that operators informed of the results of thorough investigations 'would readily identify hazards in their operations that were similar to those contributing to explosions in other facilities, e.g., propane leaks during drier installation. Operators also tend to assume that their employees are aware of hazards even though they themselves may have only a slight understanding of the mechanisms of grain elevator explosions. Many employees interviewed following the explosions investigated by the pane] showed amazing ignorance about how dust explosions occur. Insufficient training of new or temporary employees in safety procedures and fire and explosion hazards can lead to a false sense of security. Lack of standards and procedures for fighting fires in elevators also has led to explosions.* Many mid} and elevator operators use outside contractors for a variety of functions such as welding and equipment installation. If the operator, as the person responsible for facility safety, does not acquaint the contractors and their employees with the fire and explosion hazards, he is endangering the facility and its employees as well as the contract workers. Ample evidence for this is given by the high percentage of explosions stemming from welding operations (see Appendix B}. Personnel. It is conceivable that an explosion can occur in a grain elevator without an overt act by some person; however, in almost every case, someone, generally an employee, does something that contributes directly to the occurrence of the explosion. It is highly unlikely that these actions are deliberate attempts to cause an explosion; quite the contrary, the personnel probably were unaware that they were-in any way responsible for a disaster. Government Interaction The establishment and enforcement of environmental air pollution control regulations in response to the Clean Air Act of 1970 may have had an effect on the dust explosion hazard. The management of grain-handling facilities responded to these regulations in a variety of ways, some of which may have increased the probability of explosions. Prior to the establishment of these regulations some of the airborne dust escaped to the outside ambient air 'through windows, doors, cyclone collectors, and other openings. Following the establishment of these regulations dust could not be allowed to escape from the structure and it became necessary to install * One of the explosions the panel investigated resulted when firemen created a dust cloud while attempting to extinguish a fire in a smoldering dust pile. In another investigation, the fire chief of a large metropolitan area admitted that his fire service did not know how to fight fires in elevators.

OCR for page 13
27 dust collection systems to keep the air In the work space tolerably clean. Some managers viewed dust collection systems primarily as a means for meeting ambient air quality standards and only secondarily as a means for reducing the explosion hazard. It should be noted that Environmental Protection Agency regulations apply only to new elevators (1978) with a capacity of 2.5 million bushels or more but that these regulations have been applied by some states to existing facilities without regard to the federal limits on size or age. In older elevators, especially country elevators, little or no attention was given to dust control in the original design. Consequently, any dust collection system added was installed under retrofit conditions and most were installed by local sheet metal companies without reference to the engineering principles of dust control. The whole subject of dust control versus government regulation has oscillated back and forth in the past 60 to 80 years. Originally some state bureaus of weights and measures prohibited any collection of dust on the assumption that the receiver would be ~hort-weighing the received grain. This restriction eventually was lifted but was partially re-instituted by one state (Iowa) in 1979 and is being considered in another (Nebraska). The purported reason for reintroducing the restriction is that dust collection leads to or enhances dust explosions. . Econanlcs It is not unexpected that economic factors influence the incidence of dust explosions. Under ideal dust control conditions, all dust, wherever generated, would be collected and only dust-free grain would be delivered from an elevator to the next receiver regardless of the amount of dust in the grain when originally received. In practical situations, however, it is very unlikely that this could happen. First, the costs, both capital and operating, of cleaning the grain are not negligible. Second, the cold material has a sale value of only a fraction of that of grain. Even the removal of only the airborne dust involves high capital and operating costs and represents a loss in salable product. Returning the dust to the grain when it leaves the elevator permits the full grain value for the dust to be received and is one method used to reduce these costs. This, of course, i Creases the amount of dust to the receiver and thereby increases his hazard especially for the case of export elevators where the grain may have passed through as many as 3 or 4 elevators. Even customers of grain elevators--those whose grain is processed, stored, or transferred from one transportation mode to another by means of an e~levator--indirectly influence the probability of dust explosions through economic pressure. Users, including mills, are the source of many economic pressures on elevator management that in turn cause housekeeping to be ignored, machinery to be operated beyond its performance limits, welding to be done while operations are under way, and employees to be overworked. Additionally, customers always seek the lowest price for elevator services and thereby influence elevator owners and operators to cut corners and not implement all the actions that could prevent explosions. However, customers should also realize that they too~bear the cost of explosions.

OCR for page 13
28 Insurance Segments of the insurance industry.have obstructed solution of the explosion problem by promulgating recommendations and requirements that have little or no real relationship to ' the dust explosion hazard. For example, although devices such as motion switches on elevator leg belts' end . thermocouples on bearings are useful in certain applications as control measures for ignition sources, they are not the substitute for adequate.' housekeeping that some insurance-organizations have' implied. Miscellaneous One factor over which there is no control is the.effect of weather on a season's grain crop. Certain growing and harvesting conditions and certain grain types produce grain that is more susceptible to breakage in handling, and this results in a higher than normal amount of dust. For. example, the harvesting of wet corn and subsequent forced drying at the elevator in place of natural drying results in ' increased kernel breakage. . Finally, specific information on the causes of actual explosions and actions that can be taken to reduce the hazards has not been made available to mast grain facility managers. This in.itself is a contribution to the explosion problem. For example, five symposia devoted almost entirely' to discussion of elevator explosion problems were..held between October 1977 and October 1979, but the proceedings of these meetings (Grain Elevator' and Processing Society'1977, National Academy of Sciences 1978, National Grain Feed Association 1979, Texas Agricultural Extension Service et al. 1978, U.S. Department of Agriculture et al. 1979b) have not been translated into user terminology and widely distributed. . REFERENCES Bartknecht, W., Explosions, cause prevention protection. Springer-Verlag, Berlin, 1981. Development Planning and Research Associates, Inc., Preliminary Benefit Analysis--Cancellation of Carbon Tetrachloride in Fumigants for Stored Grain, for the U.S. Environmental Protection Agency, p. viii-18, April 1980. - Federal Grain Inspection Service, Policy and Procedures Upon Encountering "Unduly Lazarus Conditions. in Grain Elevators, Instruction 370-3, Revision 2, FGIS, Washington, D.C., 1980. General Accounting Office, Grain Dust EKplosions--An Unsolved Problem, GAO, Washington, D.C., 1979. i

OCR for page 13
29 Grain ~ evator and Processing Society, Proceedings of the International Symposium on Grain Dust Explosions, Minneapolis, Minnesota, 1977. Hall, Leland L., Insurance industry's view, in Proceedings of the International Symposium on Grain Elevator Explosions, National Academy of Sciences, Washington, D.C., 1978. National Academy of Sciences, Proceedings of the International Symposium on Grain Elevator Explosions, NAS, Washington, D.C., 1978. National Fire Protection Association, Guide for Explosion Venting, NE7PA No. 68, NFPA, 80s ton, Massachusetts, 1978. : National Grain and Feed Association, Proceedings of the Elevator Design Conference, National Grain and Eyed Association, Fansas City, Missouri, 1979. National Materials Advisory Board, The Investigation of Grain Elevator Explosions, Report GRAB 367-1, National Academy of Sciences, Washington, D.C., 1980. Pa finer, K.N., Dust Explosions and Fires, Chapman and Hall Ltd., London, 1973. Rowe, W.D., An Anatomy of Risk, John Wiley and Sons., New York City, - p. 28-33, 1977. T. E. Stivers Organization, Tnc., An Economic Impact Study of Proposed OSHA Electrical Standards in the Grain Elevator Industry, for National Grain and Feed Association, . p. 37, May 1980. Texas Agriculture Extension Service, Texas A&M University and Texas Grain and Feed Association, Causes and Prevention of Grain Elevator Fires and Explosions, Texas ASM University, College Station, 1978. U.S. Bureau of Mines, Explosibility of Agricultural Dusts, Report of Investigation 5753, USBM, Washington, D.C., 1961. U.S. Department of Agriculture, Handbook of Agricultural Charts, Handbook 561, USDA, Washington, D.C., 1979a. U.S. Department of Agriculture, National Grain and Feed Association, and Ransas State University, Proceedings of the International Symposium on ~ Grain Dust, Manhattan, Kansas, 1979b. U.S. Department of Agriculture, Prevention of Dust Explosions in Grain Elevators--An Achievable Goal, USDA, Washington, D.C., 1980.

OCR for page 13
30 \ 1 Verkade, M. and C:hiotti, P., Literature Survey of Dust Explosions in Grain-Mandling Facilities; Causes and Prevention, Energy and Mineral Resources Research Institute, Iota State University, ones, March 25, 1976. Wade, F.J., Hawk, A.~. and Watson, C.A., Investigations to Determine Dust Concentrations and Characteristics Inside Enclosed Work Areas and Equipment at Large Grain Terminals, final report, USDA/Cargill Cooperative Agreement No. 1-2-2 5-A-29S7, 1980 . l 1 .,