The Severn Tunnel/Chapter 3
THE COMMENCEMENT OF THE WORKS UNDER THE CONTRACT.
Commencement of the works—1880Immediately after the signing of the contract, arrangements were made to commence building the new engine-houses and erect the large engines which had been purchased by the Company. It was estimated that it would take rather more than six months to complete the erection of these engines and of the pumps connected with them. The work to be done was the erection of an engine-house for the 75-inch engine, with a boiler-house containing six Cornish boilers; a separate engine-house for the 70-inch engine, with a boiler-house for seven Cornish boilers of rather smaller dimensions.
The engines and pumps were made by Messrs. Harvey and Co., of Hayle, in Cornwall, and they were not able to promise delivery of the machinery in less than four or five months. It was thought, therefore, that the works would continue in their then state of desolation up to midsummer, 1880, except for the work to be done in erecting these houses and the machinery. Commencement of the works—1880. On obtaining, however, full particulars of the manner in which the headings had been worked from the Old Pit, especially the heading running westwards into which the Great Spring had burst, I thought that it would be quite possible to stop the mouth of the heading and the flow of water from the Great Spring by lowering into the shaft two shields made to fit, as nearly as possible, to the interior of the shaft, and strutted across the shaft in a secure manner. The difficulty was that the mouth of the heading was then 140 feet below the surface of the water; but, having obtained the sanction of Sir John Hawkshaw to make the attempt, my plans were made, and as early as the 6th January, eighteen days after the contract was signed, operations were commenced.
According to the best information we were able to obtain from those who had been in charge of the works, the quantity of water running into the heading under the river was not more than 2,000 gallons per minute, and to deal with this we had, in the Iron Pit, two 26-inch plunger-pumps, each capable of lifting 2,500 gallons per minute, and in the Old Pit an 18-inch plunger-pump capable of lifting 1,200 gallons per minute; so that if we could by any means succeed in stopping back the Great Spring, we had ample power to clear the rest of the works of water.
The construction of the shield and the method of securing it in the shaft are shown upon the accompanying plate.
Commencement of the works—1880. The shields were of oak, made to fit as nearly as possible to the interior of the shaft. One shield, when put together, was lowered to cover the entrance to the western heading; then the second in like manner lowered to cover the entrance to the eastern heading; and when lowered to their proper position, heavy struts, also of oak, were to be fixed between them and wedged up tight.
In addition to this the faces of the shields, where they were to come in contact with the brickwork, were padded with soft material soaked in tar.
The only difficulty about the operation was that the shields must be lowered by signals received from divers, who must be down the shaft at the level of the headings, and that the struts must be fixed and wedged up also by divers at the same level.
The depth of water from the shield to the surface being about 140 feet, the pressure upon the divers was so great that very few men were able to bear it at all, and no man could do work requiring great physical exertion under that pressure. In order to reduce the pressure to some extent Sir John Hawkshaw consented to my starting the pumps and lowering the water in the shaft 50 feet. The three pumps were accordingly started on the 6th January, and the lowering of the shields commenced immediately afterwards.On the 10th January, No. 2 26-inch pump broke down; the top valve, not being properly secured and Commencement of the works—1880.
held down, jumped out of its seat and would not work.
Arrangements were made at once to lower in the Old Pit a 15-inch pump to help to hold the water down till the shields were fixed.
The shield to close the western heading was made with two doors opening as flaps into the heading; the size of the opening of each door was 2 feet 3 inches by 1 foot 6 inches. On the 14th we found, to our surprise, that the top valve of No. 2 pump had, from vibration or some other cause, gone back to its seat again, and we re-started the pump; but on the 16th, to add to our troubles. No. 1 pump, which had hitherto been working well, broke down utterly, and the pump was useless.
It was impossible to attempt to repair this pump under water, and our only hope then was that we might be able, by closing the door and the sluice at the bottom of the Iron Pit, to empty the pit and then repair it. If the shaft had been properly constructed there should have been no difficulty in doing this, the shaft being tubbed with iron to within 10 feet of the bottom. Unfortunately the lower 10 feet was lined with brickwork only 18 inches thick. This was cut away on the north and south sides of the shaft to receive the iron girders and the sluice, and the brickwork had not been properly made good round these. We therefore found every effort that we made to clear the pit of water futile. More water Commencement of the works—1880. came in through the brickwork and the sluice when it was shut than one 26-inch pump would lift, i.e., more than 2,500 gallons per minute.
No. 2 pump, however, continuing to work fairly well, we continued lowering the shields, and, with occasional stoppages for repairs, we completed the fixing of the shields by the 24th January. We at once found that the shield in the western heading leaked badly all round.
The flap-doors were opened, and a large number of bags full of Portland cement was passed through by the divers, and built up inside the shield for the purpose of stopping any leakage of water.
On the 29th January, this cement being all placed behind the shield, the pumps were stopped to allow it to set properly before strain was brought upon it.
A 15-inch pump was added to the 18-inch in the Old Pit, worked by the 41-inch beam-engine, and on the 5th February the pumps were again started, and by the 7th the water was lowered in the Iron Pit 72 feet.
A diver was then able to examine No. 1 26-inch pump, which had broken down; and he found that the H-piece was broken.
Many plucky attempts were made by the divers, especially by Lambert, to ascertain where the water came in at the bottom of the Iron Pit, and whether anything could be done to stop the leakage, so that the one pump might empty the pit, and permit us to repair the broken H-piece. Commencement of the works—1880. On one occasion, on the 9th February, the suction of the pump drew Lambert so fast against the perforated wind-bore or suction-piece, that it required three men upon a rope to pull him away.
We were compelled, after this experience, to give up all thoughts of repairing the broken H-piece for the present.
On the 16th February, Lambert, the diver, reported that he found the door in the tubbing of the Iron Pit was not properly closed, the rubber with which it was faced being turned up on one side, and allowing a considerable quantity of water to leak through.
The pumps were stopped on the 16th February, to allow the water to rise in the pit, and Lambert opened the door, put the rubber straight, and closed it again. This was all done in two hours, and the pumps started again. On the 20th we succeeded in reducing the water in the Iron Pit to 20 feet above the bottom of the shaft, and for a short time we had strong hopes of being able to clear the pit of water and repair the broken pump; but, on the 21st, another accident happened to No. 2 pump—one of the tappets being bent. In two hours a new tappet was put on, and the pump started again.
I determined to remove two more 15-inch pumps from the Hill Pit, and fix them in the Old Pit to increase the power there; but on the 2nd March the top valve of No. 2 pump jumped out of its seat again.
Commencement of the works—1880. We then got an extra valve made to drop down inside the rising-main of the pump. It was attached to wooden pump-rods 6 inches square. When lowered nearly to the top of the original valve, it was held down in position by the rods, and the pump started on the 31st. We found that it did fair duty, but after working four days it was evident that the shield would not hold back the water sufficiently to allow the pumps we then had to deal with it, and the attempt to pump out the shaft was abandoned till the new 75-inch engine with the 38-inch bucket-pump was ready for work.
In the meantime Sir John Hawkshaw had decided in January to lower the gradient of the tunnel at the ‘Shoots’ 15 feet, and to keep the gradient eastwards towards Bristol parallel to the old gradient at a depth of 15 feet below it, by this means preserving the gradient of 1 in 100 against the load. From the ‘Shoots’ westward the gradient was altered to 1 in 90, so that the lowering at the Old Pit at Sudbrook was 12 feet 7 inches; at the Marsh Pit, 7 feet 9 inches; and at the Hill Pit, 4 feet 9 inches; and the gradient ran out into the old levels in the open cutting on the Monmouthshire side.
Having decided to lower the gradient, and feeling thus greater security under the river, Sir John decided to allow me to commence the work at other points.A shaft, 18 feet in diameter, was commenced at Sudbrook, on the centre line, directly opposite the Commencement of the works
iron shaft, on the 11th February, and arrangements were made for sinking a pumping-shaft on the Gloucestershire side at Sea-Wall, which was to be 45 feet to the south of the existing pit, and kept as a separate pumping-pit.
It was also arranged to sink two shafts at a point about 26 chains from the Sudbrook shafts westwards, to commence the work to the west of the point where the Great Spring had broken in, and it was at that time thought possible that we might, after these shafts were sunk, arrange to pump all the water from the Great Spring at that point, and keep the rest of the work dry.
The shaft on the Gloucestershire side, known as the Sea-Wall Shaft, was to be sunk 27 feet below the old levels to allow for the lowering of the gradient 15 feet, and for sufficient sump-room to hold the water for the pumps. The sinking of the shaft was commenced on the 18th March, and was proceeded with, with a little difficulty, till the level of the heading was reached. A cross-heading was driven from the new pit to the old one, 10 feet above the bottom of the old pumps, and what water was found in the new pit was allowed to flow through this to the pumps.
The sinking below that level was much more difficult. There were difficulties with boilers, and difficulties with pumps, and it was not till the 17th July that the shaft was finally completed, and the pumps fixed. The brickwork of the shaft had Commencement of the works—1880. been completed on the 7th, and ten days had been occupied in removing the pumps from the old pit to the new.
A heading had been driven eastwards before the contract was let to me, for a distance of 1,020 feet from the Sea-Wall Shaft. Some time in the summer of 1879, the roof at the end of this heading broke through the marl into a bed of gravel. There had been a great run of gravel and water, and the heading was partly filled. A head-wall had been built across the heading, 240 feet from the shaft, with a door. The door had then been closed, and the water from the gravel shut out. Before the new shaft was commenced, this door was opened, and the heading east of it examined. It was found to be in a bad state, supported by very small timbers, and not at all safe. Some extra timber was put in, and the door, which had been fixed by the Company in this heading, was again closed.
Another cross-heading was driven to suit the new gradients after the pumps were fixed at Sea-Wall, and a commencement was made at lowering the heading westwards to the altered levels.
To secure the ground, a small piece of the tunnel was commenced in a peculiar way, which was afterwards largely adopted.The lowering of the gradient to so great an extent as 15 feet made the existing headings useless for drainage purposes, and it was decided to put in the arch of the tunnel down to springing Commencement of the works—1880.
only, and after the headings had been lowered, to put in the invert and underpin up to the arch. This was carried out through more than a mile of the work, and everywhere successfully.
The new winding-pit at Sudbrook was commenced on the 11th February, and was sunk to a depth of 30 feet by the 8th March. The shaft was 21 feet in diameter outside the brickwork, and 18 feet inside. It was impossible to sink this shaft more than 40 feet till the water was pumped out of the adjoining works. A bore-hole was started on the 20th April to allow the water to drain out of the pit into the heading below, when the new pumping-plant should have cleared the other works of water.
The two new pits at 5 miles 4 chains, known as pumping and winding shafts, were commenced on the 23rd February; and on the same day a long culvert, to convey the water out to the edge of the river, was also commenced. These two shafts reached a depth of 30 feet on the 8th March. On the 2ist April the winding-pit was 63 feet deep, and the pumping-pit 65 feet. On the 2nd June the winding-pit reached a bed of very hard conglomerate rock at a depth of 96 feet. This rock, which was known to exist, and had been found in the Sudbrook Pit at nearly the same level, had been nowhere met with in a greater thickness than 9 feet, and in many places it was known to be only 3 feet thick. In this unfortunate shaft the thickness proved to be 26 feet. The rock was very jointy, and full of Commencement of the works—1880. fissures, yielding immense quantities of water, the water spouting through the fissures under a head of about 100 feet; and it was not till the 18th July that the shaft reached the bottom side of the bed, and entered the fire-clay shale, which proved to be perfectly dry. The bottom of the tunnel was reached at the end of July, and profiting by the experience which we had obtained at Sea-Wall, we drove a cross-heading 4 feet 3 inches below the level of the invert of the tunnel in the direction of the pumping-pit. From this heading we sunk the bottom part of the pumping-pit in shale, which was perfectly dry till we reached within a few inches of the bottom, where a small spring was met with, which evidently was under a pressure of at least 100 feet.
By putting in a small iron pipe to allow this water to rise into the cross-heading, we were able to complete the brickwork at the bottom of the pumping-pit before we allowed the water from the conglomerate above to come down upon it. When this brickwork was completed and the cement properly set, a bore-hole was put in down the centre of the shaft, and the water allowed to flow into the bottom of the pumping-pit and through the cross-heading to pumps fixed in the line of the tunnel. The shaft was then completed, both as to sinking and brickwork. The pumping during the sinking of these pits had been done by a variety of pumps, but finally by a 15-inch bucket-pump.
While the shafts were being sunk a large Commencement of the works—1880. engine-house had been built opposite the pumping-pit, and in it had been fixed a 70-inch Cornish beam-engine. A horizontal engine had also been fixed at the pumping-pit with two 15-inch bucket-pumps, and these pumps started work on the 1st November.
The fixing of the larger pumps in the pit, which were 28-inch bucket-pumps, was also pushed forward, and the engine with the one 28-inch bucket-pump was started on the 15th November, and continued steadily to work till the 18th December; by which date the second 28-inch pump had been completed, and the engine was then started working the two 28-inch pumps.
On the 31st May permission was given to start the tunnel-works on the Sea-Wall side of the river, and the mining for the shaft-length was commenced on the 22nd June.
As there was a prospect now of getting a strong force to work at various points, it became necessary to arrange the dwelling-houses for the men.
On the 9th March I arranged with the owner of the land, Mr. C. E. Lewis, of St. Pierre, to take on lease for a period of six years a plot of land near the main shaft at Sudbrook, on which to erect cottages. I had also for some time been experimenting with the various clays and marls in the neighbourhood, with a view to making bricks, and, the experiments proving successful, I decided early in the year to construct a brickyard of considerable extent near the 5 mile 4 chains pit. On account of the season it Commencement of the works—1880. was necessary to provide for drying the bricks, as they were made, under cover, and we commenced the arrangement of the brickyard and the erection of the brick-drying shed on the 23rd February.
A plan is given showing the state of the brickyard and the houses that had been built at the end of 1880.
The bricks that were made from the marsh clay, with a small addition of sandy marl, proved to be of excellent quality, and the buildings, after six years, are as perfect as when first erected.
When the shaft at 5 mile 4 chains was sunk through the conglomerate, just above the level of the springing of the arch of the tunnel, I noticed that the material which was raised from the shaft was very similar to that used in the Cattybrook Brick-yard for making the vitrified bricks. It at once struck me that the heading into which the water had broken between Sudbrook and 5 miles 4 chains must be wholly in this fire-clay shale.This was stoutly denied by the men who had been employed upon the works, but it proved to be the case, and I determined at once to endeavour to make a considerable quantity of vitrified bricks from this fire-clay shale upon the ground. This led to a large increase of the brick-making plant, and finally to the erection of eight Staffordshire kilns, with a drying-shed 100 feet by 150 feet, and a very heavy crushing-mill (similar to those used in Staffordshire Commencement of the works—1880.
for the blue bricks), to crush the shale before it went to the brick-making machine.
While these works were going on the erection of the large engine-house for the 75-inch engine had been steadily progressing. The brickwork being completed, the erection of the engine itself was commenced on the 24th April.
The lowering of the 38-inch pump in the Iron Pit was commenced on the 4th June. On the 16th June this pump fouled a piece of timber in the pit, and the diver being sent down to explore, reported that there was a heavy piece of oak lashed across the pit which prevented the pump from being lowered.
On the 17th June the one pump we had still fit for work in this pit was started to relieve the pressure for the divers to commence removing this timber. It was then found that a timber-stage extended over the bottom of the pit above the sump.
Again the divers had to descend to remove this stage. By the 25th the stage was removed and the pump lowered, when we found that it was resting on a flange of the 15-inch pipe which had been placed as a column under one of the girders carrying the 26-inch plunger-pumps. Lambert again examined the bottom of the pit, and found that the flange of this column projected nearly 6 inches from the side of the girder.
The size of the pump had been fixed by Sir John Hawkshaw, upon information received from those Commencement of thw works—1880. who had been in charge of the work for the Company when the shaft was sunk, and the size fixed (38 inches) was just as large as would pass between the girder and the brickwork in the side of the shaft The projection of the flange of this column made it impossible to lower the pump into the sump of the shaft.
Attempts were made to remove the column sideways, but the diver was unable to do so. It was suggested that the upper flange should be cut or broken off. This was also beyond the power of the diver under such a depth of water, and almost disheartened by these repeated difficulties, it was determined to start the pump, and try to lower the water sufficiently to repair the broken H-piece of No. 1 26-inch plunger-pump.
Everything being ready, pumping was commenced with the 75-inch beam-engine and the 38-inch pump, and No. 2 Bull-engine with the 26-inch pump, at 7 a.m. on the 2nd July, and in 8½ hours the water was lowered to within a few feet of the bottom stage; when, at 3.30 p.m., the 38-inch pump suddenly broke all to pieces, and in a few hours the shaft was again full of water.
The breakage of a pump of such large dimensions was of itself sufficient to terrify all those who were near it. Suddenly the heavy pump-rods, losing the resistance of the water, ran out with a crash, the outer end of the beam (which weighed nearly 23 tons) striking with immense force the Commencement of the works—1880. catch-wings, provided in case of such accidents. Before the man could stop the engine it came in again with a thundering blow, and then one of the men who had been watching at the bottom of the pit ran up with terror in his face to say that the great pump had burst near the bottom, and that a piece more than 18 inches across had been driven out of the working-barrel close by him as he stood on the staging. It was useless to work the other pump, so it was stopped, and in a few hours the works at the main shaft were again in the position they had been seven months before—full of water to the level of the tide.
At first no one could understand the cause of this accident; but without wasting time we proceeded again to pull out the 38-inch pump.
The rising-main, being of wrought-iron, was found to be perfect as when it had been lowered; but when we reached the working-barrel, which was of cast-iron, it was found to be split from end to end, and a large piece broken out of the one side, very much as the man had described it. The valve-piece below the working-barrel was also split, and the valve was missing; this latter was found in the suction-piece or wind-bore.
On examination of the pump when pulled out, we found that the bottom valve had had no valve-seat on which to rest, but was let into a taper in the cast-iron valve-piece; the taper being 16 inches long, and only ⅜ ths of an inch larger at the top than at the Commencement of the works—1880. bottom. The cause of the failure of the pump was then seen. The valves, both bottom valve and top valve, were of a new patent construction, called ‘Hat-band Valves.’ They consisted of a series of steps bored with a large number of ½-inch circular holes, and each step covered with a band of rubber ¾-inch thick; the principle being that the pressure of the water through these holes should expand the rubber, and find an escape for itself as the engine made its stroke, and close again over the holes as soon as the pressure was relieved.
It became manifest that these rubber-bands had offered a very great resistance to the passage of the water; and when the water was lowered to within 10 feet of the bottom of the shaft, the column of water standing nearly 180 feet in the rising-main above the bottom valve, the moment the bucket-valve began to make its downward stroke, the water not escaping freely through the bands of rubber, increased the already enormous pressure upon the taper-seating of the bottom valve, split the valve-case, and drove the bottom valve into the wind-bore. The engine thus released, ran out with great force; the column of water in the rising-main following. Unfortunately the bucket-rods were attached to the bucket without key or nut, but by a taper in the end of the rod, identical with the taper in the bottom valve below.
The shock of the water on the return-stroke of the engine tore the rod through this taper, splitting Commencement of the works—1880. the working-barrel, and driving out the large piece spoken of before; and the pump was a wreck.
All these matters were, of course, carefully inquired into on the spot by Sir John Hawkshaw and his assistants; and, seeing at once that the accident was the result of a defect in the pump which they had undertaken to furnish to me for the work, he ordered the broken parts to be replaced by others.
The large pump was all taken out of the shaft by July 15. To appreciate the labour entailed, it must be understood that the pump-rods alone were balks of timber 45 feet long and 15 inches square, with heavy iron mountings; then the rising-main, consisting of 9-feet lengths of 40-inch wrought-iron pipes, had all to be lifted out again, unjointed and laid on one side, before we could get to the very bottom pieces which were broken.
New four-beat valves were ordered by the Great Western Railway Company from Messrs. Harvey and Co., of Hayle, with a new valve-piece and working-barrel; and a new suction-piece or wind-bore, made oval instead of round, so that it might be lowered in the space between the lining of the shaft and the projecting flange of the pipe on which the girder rested. In ordering these care was taken to see that the bottom valve had a good level seating in the valve-case, and that the pump-rods were properly attached to the bottom of the bucket. The makers were pressed to use all despatch in furnishing the new work, and by dint of great Commencement of the works—1880. exertions the pump was again ready for starting on the 12th October.
A new H-piece had also been provided for the 26-inch plunger attached to the 50-inch Bull-engine in the Iron Pit, and spare valve-pieces and valves for both the 26-inch pumps.
The new working-barrel valves and the valve-piece for the large pump were made 35 inches in diameter instead of 38 inches, and the pump was from this time spoken of as the 35-inch pump.