|
The Cause Of The ExplosionPoint of Origin and Spread of the Explosion 59. F. J. Hartwell, a Senior Principal Scientific Officer of the S.M.R.E., said that the indications of the directions of blast and flame suggested to him that the explosion started in the vicinity of O.10 intake roadhead; at this roadhead there appeared to be a zone of small disturbance in which little dust was raised but relatively prolonged burning of rich firedamp had occurred. The heavy burning observed on timber close to the ripping lip indicated slow burning at that point; because of the gradient of the road, any accumulation of firedamp would have been thickest at the face of the ripping and this view was consistent with the degree of scorching of timber seen at floor level there. Had the explosion originated anywhere but at the O.10 intake roadhead, the flame would certainly have been forced on to O.10 face and this had not occurred. Taking into account the probable concentration of the firedamp and the fact that the flame reached the floor, he estimated that the amount of pure methane present at the time of the explosion was between 150 and 200 cubic feet. This quantity would have been enough to produce the violence necessary to raise coal dust farther outbye in O.10 intake. He considered that the explosion subsequently was one mainly of coal dust, principally that raised from conveyors for it died out on the tram haulage section of the main intake and on the main return. The prolonged burning at O.18 return roadhead could only have been due to the inflammation of firedamp from open breaks, together with some from the adjacent waste. The signs of extra violence found near O.10 supply roadhead might have been due to firedamp forced out of the waste by the explosion in the intake. 60. Hartwell had carefully considered, and put forward tentatively, a possible alternative theory visualising that a preliminary ignition might have occurred at O.10 supply roadhead, that flame then passed through the waste to ignite gas in O.10 intake roadhead, and that this in turn caused the coal dust explosion. In a number of cases flame has been proved to have travelled long distances in open breaks across wastes. But to satisfy the theory in this case it was necessary to postulate either that the ignition at the supply roadhead coincided with the accumulation of firedamp at the intake roadhead, or that the ignition in the supply roadhead forced firedamp, not on to the face, but by some means into the intake roadhead, where it was virtually unmixed with air when reached by the flame. Neither of these hypotheses seemed likely. He, therefore, adhered to the opinion that the explosion started in O.10 intake roadhead. 61. With one exception, to which I shall refer later, mining experts of the National Coal Board, the National Union of Mineworkers and H.M. Inspectorate all agreed with this view. 62. There was not immediately apparent in this location the necessary coincidence of an incendive agency and a source of firedamp in quantity sufficient upon ignition to raise and ignite coal dust. Even so, careful consideration of the possibilities leads me to accept O.10 intake roadhead as the most likely place of origin of the explosion. Source of Firedamp 63. The fact that fire-damp had not been found in the O.10 intake ripping since 9th May — some seven weeks before the explosion — suggested that something unusual must have happened to cause the accumulation there of about 150 cubic feet of firedamp in a period not longer than the five hours or so which elapsed between the night shift deputy's examination at about 5.30 a.m. and the explosion. Firedamp persisted after the explosion. For example, two days later H. L. S. Johnston, Under-manager of the No. 4 Pit, observed one and a half per cent. midway up both O.10 intake and O.10 supply gate. The ventilation had not then been restored and so a large emission of firedamp is not necessary to account for Johnston's observation. The possibility arose of firedamp having migrated to the roadhead by layering along the roof of the roadway from the large cavity some 100 yards outbye on this road but, in the light of the ventilation tests made afterwards, I conclude that it was unlikely to have happened. Tests in boreholes later drilled into the roof between the cavity and the ripping did not indicate any substantial flow along breaks or bed separation channels. 64. Another possibility was that the firedamp reached the roadhead as a consequence of displacement of firedamp by a large fall of roof in the waste. In all probability any firedamp in the waste would normally be confined to the upper and return end of it and, even if there had been such a fall, firedamp would not have been forced out to the roadhead. 65. More likely than the displacement of already released firedamp is the opening of new channels for firedamp emission, perhaps from the "rider" coal, by ground movement. There are many recorded cases of large feeders of firedamp having issued quickly, but such feeders have usually been associated with obviously abnormal ground movement and have not been short-lived. The investigations after the explosion discovered neither such ground movements nor a substantial feeder. There could, however, have been enough extra roof movement to create new channels for a slow issue of firedamp close to the roadhead itself. J. Grindle, a repairer, spoke of the appearance of bed separation at the roadhead the day before the explosion and R. H. Law observed it after the shotfiring on the night shift. A possible reason for a change in conditions at the roadhead was put forward by L. G. Fear, who drew attention to the existence of a small fault in the Marine workings the line of which, if continued, would have passed through the roadhead, and to a change of strata inclination near the roadhead. Another possibility is that the pillar of coal between O.10 face and the Marine workings, having been reduced to about 55 yards in width, must have been heavily loaded and so in a condition conducive to extra roof movement at its edges. 66. I think it probable that the firedamp accumulated slowly from small feeders close to the roadhead. With an airflow of 5,000 cubic feet per minute in this twelve feet arched roadway, the velocity near the roof would have been very low and it is unlikely that, in the absence of a sheet, the air would have swept away any such accumulation. There is no reason, however, to suppose that it had reached proportions detectable on a flame lamp when the pre-shift inspection was made at 5.30 a.m. I was impressed at the Inquiry by the fact that although firedamp appeared quite frequently, it was not suggested that the deputies had ever failed to make proper examinations or to take appropriate steps when they found it. V. Handing, the deputy for O.10 unit, who was killed, may well have examined the place, probably in the beginning of his shift, and I feel sure that had he found an appreciable quantity of firedamp in the ripping he would at least have erected a brattice sheet. Even so, and allowing for slight dispersion, an issue into the ripping of little more than one cubic foot of firedamp per minute from 7.30 a.m. onwards would have been sufficient to produce the minimum of 150 cubic feet that Hartwell estimated to have been involved in the beginning of the explosion. Means of Ignition 67. The results of the investigation discount as possible causes of ignition the safety lamps and the electrical and mechanical plant used. Of the known possible causes of ignition there remain for consideration the illegal use of a naked flame for smoking, shotfiring and frictional sparking. Contraband 68. I have considered carefully the possibility implied by the finding of an unsmoked cigarette, two live matches and a plastic bag, but I am satisfied that the explosion was not caused by the striking of a match either accidentally or by someone trying to smoke. The point of ignition was at O.10 intake roadhead and the nearest body was found over 100 yards away. Shotfiring 69. Preliminary consideration of an igniting agent pointed to the possibility of a 'hanging flame' from shotfiring and L. R. James, Head of the Safety Department, National Union of Mineworkers, South Wales Area, remained of the opinion that this was the most likely cause of the ignition. The round of four shots fired in the roof ripping near the face of O.10 intake roadhead at about 3.00 a.m. on the previous night shift could conceivably have ignited a continuous feeder of gas in a bed separation or other break. Such a 'hanging flame' sometimes — fortunately only rarely — continues to burn and could cause an explosion should a mixture of air and firedamp move upon it. The omission of stemming from the back of the shotholes in O.10 intake face rippings would have increased the risk of ignition in a roof break a little beyond the end of the hole and therefore undetectable. It seems probable that bed separation evident after the explosion was present to some extent before the shots were fired and that there would have been some firedamp in the overlying beds. Plate I and Plan No. 5 both relate to this ripping as it appeared some time after the explosion when the investigation focused attention on that place; by that time the fore-poling had been advanced. 70. A search of British records of 'hanging flame' incidents disclosed reports of 24 cases since 1934. Six of these arose from shots fired in coal. In ripping shots, which are fewer in number but far more dangerous, there were seventeen recorded instances of 'hanging flame' in the 26 years. Three of these could not be closely located from the information given, but it seemed that thirteen others occurred after the firing of shots in return road heads. The only ignition which led to a disaster was also the only one which occurred in an intake road. This was at North Gawber Colliery in 1935[1]*. In this case the road however, at the rise end of a face and near a partly-closed airway leading to a face in advance of it. 71. In none of these cases did ignition occur in conditions similar to those which obtained in O.10 intake roadhead. It is difficult to visualise circumstances in which firedamp quantity sufficient to account for the explosion could, over a period of seven hours or so. have accumulated in the ripping without reaching a flame left from shotfiring, unless the flame was in a bed separation break in the lower part of the ripping. That was perhaps possible but I do not think it at all probable. Frictional Ignition 72. Many witnesses described sparks which were sometimes produced either when the wedges were knocked out of the clamps of friction props or roof bar joints, or when released props, in falling, struck supports still in position. It is well known that sparks may result in this way or from operations such as the striking of the edge of a prop with a hammer. A great deal of investigation. however, has failed to show that sparks liable to cause ignition of firedamp are produced by this kind of steel-to-steel contact. There is no history of any such ignition. 73. Sparks, or more properly frictional heatings, resulting from impact involving certain kinds of rock, however, are dangerous. Firedamp can be ignited when, in machine cutting, the picks strike quartzitic rock or pyrites. Nothing of the sort could have happened in this explosion, however, as the machines had not been cutting for some time before. 74. Firedamp may be ignited by a particular type of blow from a steel hand-pick on some kinds of rock. Seven such cases were recorded in 1930 in a Safety in Mines Research Board Paper by M. J. Burgess and R. B. Wheeler[2], who were able to obtain ignitions in the laboratory, More recently, the S.M.I.R.E. has designed an apparatus to simulate mechanically the type of blow that causes ignition with a hand-pick. Tests were made in an explosive atmosphere with samples of quartzitic sandstone that had been exposed in the roof of O.10 intake roadhead, as indicated in Plan No. 5. Four ignitions were obtained in six tests, during which the rock surfaces were struck a total of 130 blows. A specimen of rock from the roof of O.10 supply roadhead containing an intrusion comparable to that of the rock in O.10 intake roadhead was also tested and an ignition was obtained after about 200 impacts. At the time of the explosion, the only place where the quartzitic rock might have been struck with a handpick was at O.10 supply roadhead. Hartwell, in evidence, discussed the possibility that firedamp was ignited there but, for reasons already given, this seems highly unlikely. 75. The S.M.R.E. has also examined the possibility of ignition of firedamp by friction between rock and rock. The results of tests by Burgess and Wheeler were reported on in 1928[3]. Other tests of this kind have been made recently and reported on in S.M.R.E. Annual Report for 1959[4]. The apparatus used consists of a rock 'slider' which is pressed with known force against the periphery of a rotating rock wheel in an explosive atmosphere; the pressure between the surfaces is measured and the time between the application of the load and any ignition of firedamp is taken. Ignition has been obtained with quartzitic rock. The greater the speed of the wheel and the longer the duration of the friction, the more likely it is to occur. 76. From the experiments it is deduced that, to produce an incendive condition, a suitable rock would first have to fall a distance sufficient to gain the necessary speed and then slide some distance on another rock; the shorter the fall of the rock, the longer would have to be the slide. 77. I have carefully examined the available records of ignitions believed to be due either to the impact of rock on rock or of rock on steel; the latter must, in the nature of things, include the possibility that the incendive impact may have been between rocks. The subject is of such importance that I have summarised the records in Appendix III. Four of the explosions referred to occurred in Canada. So far as this country is concerned, six of the seven instances mentioned were in South Wales. Sir Henry Walker, then Chief Inspector of Mines, in his report on the Marine Colliery disaster in 1927[5], considered that the explosion may have been due to a stone falling on stone. He cited as supporting evidence for this view possible similar incidents which had occurred between 1896 and 1927 at Maindy, Ferndale and Lletty Shenkin Collieries. In more recent years, other incidents have occurred at Cwm Colliery in 1949[6] and at Lewis Merthyr Colliery in 1956[7]. 78. It seems clear that, if quartzitic rock falls and strikes either a steel object or possibly pieces of similar rock with sufficient impact, an incendive condition may result. Hartwell suggested that the impact of the mechanised pick used in the experiments he described would have been about equivalent to that of rock weighing 260 lbs. falling for a distance of three and a half feet onto a steel object. He thought that the most probable cause of the explosion was that there was a fall of rock bringing down with it firedamp that had accumulated at the ripping and that one of the larger pieces of rock struck the canopy in such a way as to produce an incendive condition and so cause inflammation of the surrounding atmosphere. Pieces of quartzitic rock, the largest being estimated to weigh about 240 lbs. were, in fact, found on the ground and on the canopy under O.10 intake ripping (Plate I). Some of it could have fallen a distance of six feet. 79. The mechanics of firedamp being brought down from near the roof by falling stone and then mixing with air have been demonstrated with perspex models, made at the Buxton Station of the S.M.R.E., of the roadheads at which the explosions occurred at Lewis Merthyr Colliery and at Sutton Colliery in 1957[8]. I commend to all mining engineers the film which has been made to illustrate these demonstrations. 80. The evidence was not completely satisfying but I incline to the view that firedamp was brought down a fall of roof and was ignited by frictional heat at the point of impact between a quartzitic stone and the steel girder forming the top of one side of the canopy. I refer later to methods of minimising the risks associated with falls of hard rock from a high cavity and from near the face of high rippings. 81. If a fall of stone of a not uncommon nature for a distance of six feet or so may be dangerous, the question immediately raised is the degree of risk involved in the routine collapse of roof in wastes. There is, however, no recorded experience of ignition of firedamp from this cause in longwall wastes in this country. This may well be because, for ignition to occur, there must be the remote coincidence of a number of conditions including the fall of a certain kind of rock, the right type and strength of impact and the presence at or about the point of impact of a firedamp-air mixture within a relatively narrow range. 82. The following questions seem to require particular attention: whether the conditions of the tests, with frequent pick blows in the same place, resemble what can occur in practice; to what extent such impacts can be related to the impact of falling stone; whether the revolving wheel test provides reliable data upon which to found conclusions regarding falling stone; to what extent the ignition hazard is affected by the mass of the falling stone and its velocity; whether there is some peculiarity in the mines in South Wales, such as the nature of stone, which may account for the preponderance there of these ignitions. All these problems, together with others relating to frictional ignition, are the subject of research at the S.M.R.E. and there can be no doubt as to the importance of this work. A Minority View 83. A. Davies, a National Union of Mineworkers' Inspector, advanced the theory that the explosion started as a result of the firing of a ripping shot in O.18 return roadhead. which is illustrated in Plate II. The well kept records of explosive issued and returned showed that a detonator and a pound of explosive, issued to the deputy of O.18 unit on the morning of the explosion, were not recovered afterwards. Davies thought that a shot may have been fired in the right hand side of the ripping just before the explosion, that this may have ignited firedamp which had continued to burn and that the deputy and others were at the time of the explosion fetching stone dust in order to put out the flame. I saw nothing in this place to lead me to think that a shot containing as much as one pound of explosive might have been fired in the ripping. It seemed to me far more probable that the missing explosive had been used in the coal near the ribside for there were large lumps which had apparently been 'turned back' a sufficient distance from the face to make a track for the coal cutting machine. It would have been a very reasonable thing for the cutterman, who had only to square out the end of the face on this shift, first to cut enough coal at the corner for such a shot and then cut again. There was a supply of stone dust very near the ripping. But the conclusive evidence against the theory advanced was that, if an ignition had taken place in O.18 roadhead, the direction of flame would have been completely contrary to the indications found. * The numbers in square brackets refer to the Bibliography at Appendix IV.
|
![]() |
||||||||
|
Mail: Webmaster |
Back |
Home |
Copyright © 1999-2024 by The Durham Mining Museum and its contributors Registered Charity No: 1110608 Page last updated: 31 Dec 2022
|
Search |
|
|||