11th JULY 1923



extracted and adapted from the report by 

A incident occurred about 6.26 p.m. on the 11th July at London Bridge on the Brighton Section of the Southern Railway. In this case after the 6.2 p.m. electric train from Victoria had arrived at London Bridge, and was waiting at No.10 platform road rear to make its return journey, an explosion occurred in the motorman’s compartment of one of the intermediate motor coaches.

As a result of the explosion two of the windows on opposite sides of the compartment in question were shattered, and also one of the windows of a passenger train standing alongside it in No.11 platform road. Six of the passengers in this train were injured by the broken glass. The other two windows of the motorman’s compartment were jarred from the closed to the open position.

As regards the stock, practically no damage was done beyond the breakage of this glass and a slight bulging in the door enclosing in the door enclosing the high tension chamber in the motor coach in which the explosion took place.

The train consisted of the following stock, marshalled in the order shown from the terminal buffer stops t which it was standing:- 

Driving Trailer 4068, Motor Coach 3212, Driving Trailer 4060, Motor Coach 3224, Driving trailer 3226 & Motor Coach 3202.

The total length of this train was 345ft. 9in. over buffers. The compartment in which the explosion took place was t the end of coach 3224 remote from the buffer stops, that is to say, at the far end of the fourth coach from the terminus.

The evening in question was a fine and very hot.

The train which included this motor coach, was taken over by motorman B.H. Smith, at the Peckham Depot, which it left a4.17 p.m., to run empty to London Bridge. Smith stated that he examined all fuses; opened the high tension chamber doors of each motor coach, and saw that everything was in good order. He examined the level of the oil in all the-oil switches and found it to be correct. He also tested the brakes and the working of the bows of all motor coaches and found them to be in good working order. He then drove the empty train to London Bridge and found that it behaved in a satisfactory manner. On arrival at London 
Bridge the two leading coaches were uncoupled, and the remaining six conches were handed over to motorman C. Horsefield. 

No verbal report was given by Smith when the train was handed over, but he conveyed to Horsefield the intimation that all was correct by putting up his hand.

Horsefield than worked two double trips between London Bridge and Victoria with this train, and it appeared to be running in the usual way. Upon his arrival at London Bridge from Victoria at 6.26, on the last of these trips, he proceeded to reverse his bows and to remove the reverser key from the controller in trailer coach No. 4068. The operation of reversing the bows proceeded satisfactorily and when Horsefield left this driving compartment they were al in the correct position for leaving the station in the opposite direction. Horsefield then pulled up the window of his compartment on the driving side and had just begun to walk down the platform when he saw a puff of smoke come from the overhead conductor near the middle motor coach. 

He continued to walk down the train find had walked past two more conches when he heard an explosion, which seemed to come from the same motor coach. He thought that there was an interval of some 20 to 30 seconds between these two occurrences. 

On arriving at motor coach No. 3224, one of the brake doors of which had been opened hwent in and found that there was still gooof smoke. He also noticed a little light under the bottom of the high tension chamber door, which looked at though something had fallen and was burning on the floor. He then went to the front of the train to lower the bows. Having done so, he I returned towards motor coach 3324, and noticed that the bow on this coach had not dropped, although the two end ones hadone so. Hthan went back to the far end of the train and told the Platform Inspector to arrange for the current to be taken off the line. Immediately afterwards he met the switch attendant, who told him that this had a1ready been done, and on his return to motor coach 3224 hfound that the bow had dropped. He opened the high tension chamber door in this coach and observed the condition of the gear inside. He then isolated the coach by taking out the jumpers and arranged for the train to be driven to Peckham on the other two motor coaches.

Platform Inspector  Milton was on duty on this platform when the accident took place. He noticed that as the bows were being reversed and as soon as the centre bow reached the overhead wire, a puff of smoke occurred at the point of contact. He then walked towards this motor coach, and just before he reached it an explosion occurred, blowing out a door light on the platform side of the motorman’s compartment and scattering the glass on the platform. This explosion he said, occurred some seconds after the puff of smoke from the trolley wire. The guard who was walking along the platform with Inspector Milton then opened the doors of this compartment on the platform-side. After the arrival of the motorman and his subsequent action already related in Horsefield’s evidence, Inspector Milton noticed the bows at both ends of the train lowered, and also observed that the centre bow remained in position. A second explosion then took place, and shortly afterwards the motorman again came up Inspector Milton said that the second explosion occurred some time after the two end bows had hen moved from the trolley wire. In his opinion, the noise of the second explosion was at least as loud aif not louder than, that of the first.. Motorman Horsefield however head nothing of this second explosion. Inspector Milton also noticed the light showing under the bottom of thdoor of the high tension chamber after the first explosion.

Guard Hall of the train confirmed  the evidence given by motorman Horsefield and Inspector Milton in regard to this sequence of events, but he did not see the initial puff of smoke when the bows were first changed nor did he notice any light under the door of the high tension chamber. He gave it, as his opinion that there was an interval of two or three minutes between the two explosions, but did not think that the second explosion was as loud as the first, nor did he notice any smoke from the second explosion, although there was a considerable quantity from the first.

Guard Cresswell, who was in charge of the train loading up at No.11 platform (that is to say, along side the train on which was due explosion occurred), and which was due away at 6.32 p.m., noticed a few minutes before his train \was due to leave a flash at the “T”  joint between the feeder and the wire bridging the two overhead conductors running along 10 and 11 platform roads respectively. He saw nothing of any other occurrence nor did he hear any explosion.

Station foreman Forster, who was crossing the yard quite close to the far end of the platform, heard, when he was between 8 and platforms a faint report, to whicat the moment he attached no importance. Hheard...nothing of any subsequent explosion.
Acting static sub station attendant F. Garman, who was on duty at Peckham Junction Central Switch cabin at the time, gave evidence that, at 6.26 pm. S.3 switch tripped. He then inserted a 6 ampere time limit fuse and closed the switch again after the regulation interval of 20 seconds. The switch then held. At 6 .27 p m .he received a telephone message from 
assistant lineman Tully, at London Bridge Yard Switch Cabin. that. T.1 trolly switch had tripped. Garman then instructed Tully to change his fuse anto close so T.I switch. At 6.28 p.m.’S.3 switch again tripped. Garman then told Tully to open and subequently to earth T.1 trolley switch. Having heard by telephone from the station staff that the other T.1 switch (that is the one in the terminal switch cabin) he also opened. Garman then closed S.3 switch in order to energise the remainder of the group section, leaving Nos. 10 and 11 roads isolated. At 6.41 p.m., after Tully had reported that the overhead line was in order, instructions were given for T.switches to become again. Tully’s report confirmed this evidence, but contained the additional information that soon as he closed his T.I switch on Garrnan’s instructions it immediately opened again. 



From the foregoing evidence, it is clear that three distinct short circuits to earth occurred between 6.26 and 6.28 p.m.; the first resulting in the opening of S3 circuit breaker, set instantaneously; the second opening both T.I. breakers in parallel set 320 amperes each, with a time log of 11 seconds; and the third opening both S3 switch, set instantaneously, and one of the T.I switches set with the normal time lag.

There are very clearly indication of a short circuit in the oil switch an no evidence whatever of short circuiting elsewhere, either upon the power or earth side of the switch.

In view of this negative evidence, together with the fact that the main transformer fuses, rated at only 80 amperes, were fond intact, it is a safe dedication to assume that  the trouble originated in and was confined to the main H,T, oil switch. The possibilities can be further restricted by the fact that the only sign of arcing and break down are to be found in contacts, which is not live except when the switch is closed. It therefore follows that all three circuits took place with the main switch closed. 

Further from the indications of heavy fusing on the short switch contacts, it is evident that when this switch did open it was carrying a considerable short circuit current; which suggests that the fusing of the contacts occurred when the overload relay fuse blew under the overload and so interrupted the main switch holding circuit. It must therefore be concluded that this overload relay fuse did not blow until the third and final short circuit, since otherwise the main switch would not have again picked up, and contact at which the fault was manifest would not then have been live. The only alternative to this conclusion 
would be the assumption that the fusing of the contacts was so heavy as to cause the switch to stick up against the gravity pull off. This pull is, however heavy 25lbs. and, moreover, had this occurred it is highly probable that the contacts would have required mechanical separation and that the switch would not have been, as it was, found open after the accident. Finally, all three short circuits must have occurred with both S3 and one at least of the two T.I line switches closed.


As is perhaps inevitable in a case of this kind, there must be an element of conjecture as to the precise nature and sequence of the manifestations. I have, after very careful consideration, come to the conclusion that the facts were substantially as follows:-

The earth fault originated in the cracked insulator of the main H.T. switch. Of the nature and probable cause of this fracture more will be said latter. It is highly probable that one or more of these cracks had developed before and possibly several days before the train arrived at London Bridge, and that a more or less limited discharge had been taking place through these cracks from the surface of the insulator to the casing of the switch. The burning of the wood lining and the grooving of the insulator indicate that this discharge had followed indifferently a path along the surface of the oil up the side of the lining to earth, or up the surface of the insulator and to earth by way of the lid. This discharge took at first the form of a light resistance leak and originated  at the bare end of the contact, running thence between the insulator and the litharge packing, in which there was no sign of puncture. The effect of this discharge was to carbonise and vaporise the oil, which had the two fold result of lowering the oil level, and by producing a heavily carbonised vapour and sooty deposit, increasing the conductivity of the discharge path. 

Conditions were therefore ripe for a more definite manifestation by the time the train arrived at London Bridge. The first of these occurred as the bows were changed, and the probable surge caused by the chattering of the forward and reverse bows when this was done resulted in the first definite short circuit, manifested by the puff of smoke, etc., seen by the witnesses, and resulting in the opening of 8.3 switch. This did not blow the overload relay switch, mainly, I think, because of its time lag. The time lag also prevented T.1, T.1switches from opening. This short-circuit resulted in further and more serious vaporisation of the 
oil, and the charging of the high tension chamber though the oil switch vents with an explosive mixture, whose diffusion may well, I think, have been assisted by the very high temperature prevailing. It is probable that this short circuit caused sufficient fusing at the contact between the overhead conductor and the bow to result in the latter sticking up when later, the motorman actuated the lowering mechanism.

Twenty seconds later, viz..at 6.26.20, the high tension circuit was again made at S.3 switch, resulting in the closing of the oil switch followed by n discharge of such intensity as to fire the explosive mixture already generated. I think that the  result of this second short circuit and explosion was so to lower the oil level as to expose the bare ends of the fixed contacts. The effect on the line was to open T.1, T.1 switches, S.3 having now its time lap fuse inserted. It is a little difficult to account for the assumed failure of the overload relay fuse ta blow on this occasion, since its capacity is considerably below that of the two T.l switches in parallel. But I think it may be explained by the supposition that the two latter did not blow exactly simultaneously together, possibly, with the effect of magnetic leakage in the overload relay current transformer.

The third and final short-circuit took place when T.1 switch was again closed. It was this discharge which was, I think, responsible for the burning of No. 3 fixed contact, now assumed to be above the oil level, and the information of the crater in the container. It was probably a more or less complete arc through the heavily carbonised oil vapour already formed, and may well have resulted in the heaviest current to earth of the three short-circuits. At any rate, it had some characteristics of a very heavy current in that it appears not. only to have operated three cut-outs, having varying constants. which were, in effect. in series (S.3 switch one of the T.1 switches and the overload relay fuse) but also to have arced heavily tit the contacts of the oil switch as it opened. It was, I think the noise of this arcing which was the chief factor in t h e second “ explosion,” which in spite of Inspector Milton’s evidence, appears, on the whole, to the been less noisy than the first. This view is supported by the evidence of guard Hall and by the fact that neither Horsefield nor station foreman Forster heard more than the one explosion.


The record of these electric trains has been a remarkably good one this case being the only electrical train equipment accident reported since their operation started in 1909. Past performance of the apparatus therefore suggests that there is little if any warrant for adverse criticism as a result of this somewhat difficult and in many ways remarkable use. At the same time, if there is any tendency to weakness as a result of the combination of insulator design with the packing compound used and the foregoing conclusion is based on the possibility that, there maybe it is more likely to eventuate with an insulator of the design concerned than with the other and older type. I therefore suggest, and I understand that the Company’s Office are in agreement with this suggestion that insulators of the ”South London” type should, as opportunity offers be replaced by those of the  “Crystal Palace” type particular care being taken that the packing is not brought down into the recess. It would also, I think be as well to modify the oil gauge so as to make the level of the oil, indicated rather more conspicuous and easily read than is now perhaps the case. I  suggest, too. that, as the circumstances of this case definiteIy indicate the liability to formation of an inflammable gas, which is lighter than air, under conditions of short-circuit, the Company should consider the possibility and advisability of providing some ventilation near the, top of the high tension chamber as a precaution against the formation of an explosive mixture therein.



(a) The accident was due to the ignition of an explosive mixture formed by the vaporisation of the vaporisation of oil in the high tension switch of motor coach 3224, caused by discharges to earth inside this switch.

(b)These discharge resulted from cracks in one of the four porcelain insulators in this switch.

(c) These cracks were due to mechanical strain on the interior of the insulator access, set up by the litharge packing, and probably aggravated by the high temperature.


( a ) That the Company should replace the deep recess type of insulator by the shallow recess type in altheir electric stock. 

(b) That the switch oil gauge should be modified to make the oil level more easily read

(c) That the question of ventilation of the high tension chamber should receive consideration. 

Make a free website with Yola