V. THE PLACE OF ORIGIN AND CAUSE OF THE EXPLOSION.

(1) The Place.

There is no doubt that the explosion occurred among the old pillars of the Main Band seam immediately below and in the vicinity of the Countess Bannock Front Drift. Whether it occurred outbye or inbye of the already frequently referred to semi-circle of stoppings designated A.1, etc., J.1, is a matter for conjecture and argument.

Unfortunately, information regarding the effect of the explosion on the New Front stopping just inbye of the airlock is wanting because the fire inbye precluded access and observation. The bricks from the hole in the left-hand partition wall of the airlock appear to have been blown outwards. Similarly the hole in the roof of the Main Band seam into the Countess Bannock Drift seems to have been blown upwards, subject, in this instance, to the reservation that the position of this hole could not be accurately determined.

Without some explanation, perusal of the plan and section on Plate No. 3 of the Main Band roadway inbye of the new Front Stopping may lead to misunderstanding. This roadway was, in fact, closely timbered with props, bars and lagging. For clarity the wooden supports have been omitted from the drawings. Once started, therefore, a fire had plenty of readily combustible fuel on which to feed.

(2) The Cause.

For an explosion to be possible it is trite to state that an explosive medium and a means of ignition are essential. Excluding blasting explosives from the first category (although they may be included in the second), the only explosive media are, firstly, inflammable gas or gases mixed with the appropriate proportion of air to render the gas or gases explosive; and, secondly, coal dust. In this instance there is no suggestion that coal dust played any appreciable part in the propagation of the explosion and consequently there only remains the alternative of inflammable gas or gases.

Since active fire due to spontaneous combustion had existed for a long period in the vicinity of stoppings G.1 and H.1 and more recently incipient fire appeared probable in the neighbourhood of the New Front stopping, it is clear that there existed the means for the ignition of an inflammable gas should this become formed in the vicinity of the intake stoppings.

It was quite clear, however, that at no time prior to the explosion had it occurred to the minds either of officials workmen or H.M. Inspectors, who had kept closely in touch with every development from the commencement of the trouble in December, 1938, that such a contingency was likely to arise.

This was perhaps natural, since the large majority of the coal mining fraternity associate the explosion risk with the occurrence of firedamp.* Those concerned with the gob fires at William Pit appear to have belonged to that large majority.

Having this belief, the facts set out categorically below seem to be ample enough to justify that ease of mind.

(a) Throughout the period of two-and-a-half years from the first detection of carbon monoxide in the return airway near the "Humbug Doors" to the day of the explosion, firedamp had never been met with in sufficient quantity to be detected on the flame of a safety lamp in the vicinity of any of the operations involved during the building of the stoppings or subsequently, either on the intake or return side. On one occasion only 0.75 per cent. of methane was found in a sample in the general body of the air taken inbye of J.1 stopping while this stopping was being built. Reference to Plate No. 2 will show that the pillar inbye of this stopping is quite close to a fault having a downthrow west of 7 Fathoms. Further outbye, where this fault crossed the main intake at Lowca Junction, there had existed for a long time in the roof above the Junction a large cavity only partially ventilated. Examinations made from time to time in the cavity had failed to reveal the presence of firedamp. Moreover, except when work was proceeding or inspections were being made inbye, the doors in the airlock were kept closed, and similarly as regards the door in the "New Front" stopping.

(b) The Main Band pillars had been formed many years ago, so that firedamp which may have existed in the virgin coal had had ample opportunity to drain away.

(c) Experiments have shown that coal in a solid mass, from which firedamp has apparently ceased to exude, may, if suddenly reduced to a fine powder, emit considerable volumes of firedamp which have been held in the interstices of the coal. Acceptance of this fact does not help in this instance, because up to within not more than an hour of the explosion nothing had occurred to suggest that any of the pillars had been disintegrated or were likely so to be.

(d) Whilst there is no positive evidence as to the presence or absence of firedamp behind the intake stoppings, the circumstantial evidence seems to warrant the conclusion that there were no accumulations and that emissions, if any, were negligible.

These facts seem almost summarily to dismiss the possibility of firedamp having been the explosive medium.

Before the Inquiry opened it was known that evidence of the presence of firedamp at the locus of the explosion would probably be lacking, and that to account for the presence of an explosive medium the possibilities of artificially produced inflammable gas would have to be explored.

On this account the assistance of Mr. J. Ivon Graham, as scientific assessor, was sought. Mr. Graham's wide knowledge of the products of the combustion, the distillation and the reactions when at red heat, of coal or carbon with water, together with knowledge of the effects of these products on the composition of atmospheric air have been most helpful. I am indebted to him for many of the observations in the following discussion of the pros and cons of the suggested source of inflammable gas.

It is quite clear from Section III of this report, "Events leading up to the Explosion", that active combustion had been going on in the vicinity of the return stoppings near the Countess Bannock Back Drift for at least 15 months prior to 3rd June, 1941. It is also clear that active combustion had been going on behind the intake series of stoppings, G.1, H.1, etc., for nearly 12 months. Having regard to the facts observed and their sequence, it is not beyond the bounds of possibility that, finally, there existed a belt of burning coal stretching between these two series of stoppings. It seems, however, only remotely possible for any large volume of inflammable gas to have been a product of that type of combustion. Moreover, had it been otherwise, so that an explosive mixture did occur and was ignited, it would hardly seem likely that force and flame would have been ejected outwards past the stoppings when there was so much room for pressure release further inbye in the large area of abandoned pillars.

It is necessary, therefore, to explore the possibilities of the production of inflammable gas in the area in front of G.1 and H.1 stoppings.

The recorded facts and observations in the cutting of the New Front stopping and at the break near G.1 stopping during the last week of May, 1941, all lead to the conclusion that spontaneous heating had commenced subsequent to the completion of the New Front stopping and the closing of its door on 7th March. It is impossible accurately to fix a date, but there seems little doubt that by the end of the last week of May it was well on its way to active fire. Several suggestions were made as to the exact situation of this final heating. It is, however, a matter of little moment whether it was just here or just there. It is sufficient to conclude that it was either in or adjacent to the triangular-shaped half-pillar which was penetrated by the extension of the New Front stopping. This stopping was intended further to reduce the leakage of air to and through G.1 and its companion stoppings; events showed that its ultimate effect was merely to divert the leakage to another weak spot, namely, the half-pillar just referred to and the break in the roof over it. Admittedly, this and the other pillars in its neighbourhood were badly fractured and air tightness or any approach to it could not be expected in them.

The conditions necessary for the production of inflammable gas were thus provided.

It seems reasonably clear that active fire was proceeding throughout the day shift prior to the explosion, and possibly for some time also before that shift. Consequently, it seems also reasonable to suggest that inflammable gases, mainly C.H.4 and Carbon Monoxide, were produced, but not in any great volume, for it is inconceivable that any large quantity could have been developed without detection either on the flame of a safety lamp or more probably by the noses of such long experienced observers as Mr. Dawson and his associates.

Even in this connexion, however, although the theory of gas developed by distillation is dismissed – and also in respect of the alternative water gas theory shortly to be discussed – there is one factor which is puzzling almost to the point of bewilderment, viz., where did the noxious products of incipient combustion, and later of active combustion, find an outlet without contaminating the atmosphere in which persons were at work and in which numerous careful examinations were made by officials during the morning of 3rd June. It rather looks as if there was access, somewhat restricted perhaps, for both products of combustion and water to the open spaces behind H.1 and further afield. It has already been observed that up to the time, about 1.15 to 1.30 p.m. or so, when the manager, Mr. Dawson, and the deputy left the site, the water which at that time had been running in the cutting for half an hour or more had made no appearance in the observable portions of the roadways between the New Front stopping and G.1. The explanation of this may be that the water seeped away in the direction of stopping J.1, the inclination being in that direction, but it does not satisfactorily fit in with the description of the conditions either in the road leading to J.1 or in the road at right angles leading to H.1 and G.1. That it was expected to make its appearance somewhere towards G. 1 was clear from the evidence of Mr. Dawson, who seemed to be surprised when he could not find it.

It remains now to discuss the water-gas theory. In this connexion three separate considerations arise: firstly, the conditions necessary for the production of the inflammable gas in sufficient volume; secondly, the conditions necessary for the accumulation of the inflammable gas; and, thirdly, the means whereby an explosive mixture of the gas was ignited.

The first and second of these present no serious problem. In the production of the gas a mass of red hot coal or coke and a supply of water or steam are necessary. These were present from the time the water was turned into the cutting at 12.20 p.m. At the Inquiry an estimate was given by Professor Granville Poole, called by the owners as an expert witness, that probably a volume of 700 cubic feet of gas was required to form an explosive mixture in the open space between the air-lock and the G.1-H.1 set of stoppings. This figure was based apparently on the assumption that the air in that space was stagnant, which, of course, it was not. Making allowance for the possible error on this account, and the relation of this to the quantity of red hot coke that was available for the reaction with water, it still seems feasible that the necessary conditions existed.

For the gas to accumulate and form an explosive mixture in the space available it was only necessary for it to be produced at a rate rather greater than that at which it could be dissipated through the stoppings or adjoining strata.

These conditions prevailed, for although the total leakage through the stoppings was not considerable the passages for it through the strata were comparatively very restricted. Any influx of gas extraneous to the normal leakage current would accordingly rapidly foul the atmosphere in this limited space. An influx of extraneous gas would, of course, occur as soon as the reaction between water and red hot coal commenced.

It is, however, immaterial to determine exactly the composition of the inflammable gas seeing that, whatever its composition and provided there was enough of it and a source of ignition, an explosion could occur.

The second essential, viz, the igniting medium, requires consideration. Professor Poole suggested that the inflammable gas produced at a fire, either in the triangular pillar or in the road immediately adjacent thereto, travelled afield, was partly diluted by air, and then returned to its source to be there ignited. Whilst this suggestion may have its possibilities it does not seem convincing. It would seem more likely that the atmosphere outside the stoppings was slowly, or it may be rapidly, vitiated by the gas until it reached ignition point. Contact with flame behind the stoppings G.1 and H.1 was then made and the flame struck back with catastrophic results.

The question now naturally arises as to whether, if the explanation of it is correct, this explosion of water gas, or of some mixture of inflammable gases of a like nature, was unique in mining history or whether there have been any comparable occurrences in the past. The answer to the question is in the affirmative, except as regards the extent of workings affected and the loss of life.

To start with, at the Inquiry it was stated that on several occasions when water was set to flow through the pipes in G.1 and H.1 the flow resulted in "pops," and on occasions flashes of flame were observed to be projected to the outbye side of the stoppings. It was suggested that these flashes were caused by the explosion of gas evolved by the reaction of the water with the burning coal beyond the stoppings.

On the 23rd December, 1908, at Cakemore Colliery, Worcestershire, three persons were fatally injured and two others less seriously when dealing with a fire due to spontaneous combustion in a roadway in the Thick Coal. The occurrence is reported in the Annual Report for that year by Mr. Hugh Johnstone, late H.M. Divisional Inspector of Mines. Put shortly, the facts were that a considerable quantity of burning and glowing coal fell from the roof into water which had accumulated to a depth of two feet on the floor. Those in the vicinity were immediately enveloped in steam and hot dust.

Although the official report is silent on the point, it was generally agreed among those concerned that the observed effects suggested an explosion of gas, probably water-gas, as a contributory factor.

I am indebted for this information to Mr. J. R. Felton, H.M. Deputy Chief Inspector of Mines, who as Junior Inspector took part in the investigation of the accident.

On 15th December, 1929, in the Louisa Old Pit of The Holmside and South Moor Collieries, Limited, South Moor, Co. Durham, a fire was discovered at a disused air crossing in a main intake and haulage road in the Low Main Seam. On breaking through the air-crossing it was discovered that the fire had a very firm hold in the overlying strata which comprised the Maudlin Seam. A water supply was available and for a time it appeared to be having the desired effect. Later ground was lost. Application of the water, however, was continued, and two days afterwards when a jet was suddenly turned on the heart of the fire, which was then very intense, there was an explosion which blew out a stopping 55 yards away and opened some doors situated 400 yards from the site of the fire. Lights were put out and workmen near the fire were thrown down. No flame was observed, and no one was hurt, which seems to suggest that only a comparatively small volume of violently explosive gas was involved. Firedamp was ruled out and it is likely that the gas was the product of chemical action between water and fiercely burning carbon.

It is possible that this narrative of comparable occurrences could be extended, though with the exception of the occurrence at Cakemore I have found no record of loss of life from this cause, but the two instances quoted seem to be sufficiently convincing that explosions have occurred without firedamp as the explosive medium.

* There is a possible alternative to firedamp. viz., gases resulting from the use of oil immersed switches in connection with electric plant. The occurrence, however, of these gases is rare, and the explosion of them still rarer; moreover, there was no electric plant in the area in which the explosion took place.