The Severn Tunnel/Chapter 6
CHAPTER VI.
TUNNELLING.
Progress of the work—1881.As it is impossible to write an account of this description without continually using more or less technical terms, it will be as well here to state briefly the methods by which tunnels are constructed, and so to explain many of the terms which are unavoidably used.
Generally tunnels are adopted where it is necessary to construct a railway or a canal through a hill at a greater depth than 60 feet. The comparative cost of open cutting, or tunnelling, may be said to decide whether a tunnel should be made or not; but if a tunnel is to be adopted, an engineer of experience in these matters will select a line of route and a line of levels which will bring the tunnel into the most favourable strata for its construction.
Of course, in dealing with the Severn, a tunnel under the river, or a bridge over it, could have been constructed to complete the connection between Bristol and South Wales.
If a tunnel were adopted, the only open question Progress of the work—1881. would be, on what line and at what level it should be driven to pass through the most favourable ground, and encounter the least risk? To determine these questions, it was necessary to put down borings and ascertain the nature of the strata. The geology of the district was carefully studied by Mr. Richardson, borings were taken at many points, and the strata were supposed to lie horizontally. But in the execution of the work they were found to be much contorted, as will be seen by the longitudinal section of the tunnel.
The tunnel was adopted by the Great Western Railway Company in preference to a bridge, and the length of 4½ miles, which was originally fixed for tunnelling, was the length between the points at which the gradients rose, to 60 feet from the surface.
Tunnelling may be in rock, or in very soft or loose strata.
Tunnels may be required to be executed with great speed; or if they are short, speed may be no object. The consideration of speed will determine, to a large extent, how the tunnel should be carried out.
Forty years ago there were continual discussions as to whether the best method of constructing a tunnel of any length was by driving a top heading or a bottom heading. If the tunnel is short, and in solid rock, no heading may be necessary.
In the long tunnels through the Alps, which generally required no lining, a top heading was driven by one gang of men, widened out by other Progress of the work—1881. gangs, and lowered to the required level by other gangs.
But generally in tunnelling it is necessary to drive a heading—first, to ensure the correctness of the line of the work; and secondly, to attain sufficient speed. And without doubt a bottom heading is the best for this purpose. If the tunnel is without an invert, the heading should be driven on the level known as ‘formation level;’ that is, 2 feet below rails. If the tunnel has an invert, the heading should be driven at the level of the top of the invert, as, if it were driven at the bottom of the invert, the brickwork of each length would block the heading.
Driving the heading at the level of the top of the invert necessitates, where many lengths are going on at the same time, bridging over each length as the invert is taken out.
A heading is a small tunnel, and if timbered at all, it is timbered quite differently from the tunnel itself.
A heading 7 feet square is quite large enough for working skips or small trolleys. If it is decided to get ordinary tip-waggons into the works, the heading must be at least 9 feet square.
If the ground is in rock, only occasional timbers may be required to support the roof when stones in the top are loose, or sound, when struck, what the miners call ‘drummy.’ The top in that case may require ‘head-trees,’ which are timbers across the line of the heading, and polling-boards.
The head-trees may be let into the rock itself, or Progress of the work—1881. may be supported on short props resting on ledges of the rock, which, from the shape in which they are put in, are called ‘sprag-props,’ i.e., short spreading props. Or the head-trees may require to be supported by ‘side-trees’ the whole height of the heading; and these side-trees may, if the ground is soft in the bottom, require ‘footblocks,’ or ‘sills,’ which latter are timbers laid across the bottom of the heading from side to side.
In driving in clay or chalk, or in the shales or marls at the Severn Tunnel, it is possible to drive a kind of benching in the top, 1 foot in advance of the face, into which the head-trees are rolled and wedged in. If boards are required to support the roof, they are put in at the same time. This head-tree rests upon the ground of the face itself; then if side-trees are required two chambers are cut, one on each side, into which the side-trees are inserted; the weight in the meantime resting on the ground of the face, which has not been disturbed, or a temporary prop in the middle. When the side-trees are in place, the head-trees are wedged up from them, and the small wedges which are used in this and other places in timbering the tunnel are called ‘jacky pages’ (perhaps ‘jack up edges’).
By these means the heading can be advanced in ordinary ground at a speed varying from 12 to 20 yards per week, allowing three or four months from the commencement to get the heading well advanced, and start what are called ‘break-ups;’ .jpg)
the tunnel itself can be completed at the same speed, the number of break-ups and the distances between them being regulated with a view to this.
A ‘break-up’ is a hole driven in a slanting direction from the top of the heading to the top of the tunnel itself, or, if the ground is bad, to a height of 18 inches or 2 feet above that. When the break-up is completed to the top of the tunnel, a top-heading is commenced similar to the bottom heading, but smaller in dimensions.
It may be taken to be about 3 or 4 feet wide at the top and 5 to 6 feet wide at the bottom, and 6 feet high.
The nature of the ground will determine the length to be taken out at once; whatever is decided upon is called a ‘length.’ In very bad ground this may be only 6 or 9 feet; but generally it will not be less than 12 feet, and, if in very strong ground, 20 feet or even 24 feet.
When the top heading is completed to the required length, a benching is cut in the top of the heading at one side, sufficiently large to receive the timbers that have to carry the roof. These timbers should be of good, fresh, round larch, varying from 12 inches to 16 inches in diameter, and are known among miners as ‘bars;’ and the bars of the length are divided into ‘crown-bars’ and ‘side-bars.’
When one bar has been rolled on to the benching prepared for it, another benching is cut upon the other side of the heading, and a second bar placed Progress of the work—1881. upon that; between these bars the roof is secured by polling-boards, the thickness of the boards varying from 1 to 3 inches, according to the weight of the ground to be carried.
When the two bars are placed as above, chambers are cut down to the floor of the heading at a slight angle each way from the length, and into these chambers are inserted props, which are tightly wedged to the bars; these props are known as ‘back-props.’
Having secured the two first bars in the manner mentioned, another chamber is cut following the section of the tunnel like a shelf on the one side of the heading, into which the third crown-bar is rolled, and then secured by back-props in the same manner as the first, the fourth, and other bars following; and when sufficient crown-bars are in this way inserted, secured by the back-props, and themselves holding the polling-boards in place, they are secured by stretchers or short timbers; between these a cutting is made, just inside the back-props, to the level fixed upon for the first sill. The sill-timbers are then brought in, in two pieces. They are large square timbers, from 12 to 15 inches square, scarfed in the middle, and when placed upon the ground in the space which has been taken out for them, are jointed together with strong iron straps, called ‘glands.’ Another row of props is then inserted under the crown-bars, one under each bar resting on the sill; and the same operation is continued till .jpg)
the whole tunnel is completed to the top of the invert.
Two sills may not always be necessary, but the above description deals with ground for two sills; the working down from the upper sill to the lower one being exactly similar to that required for working to the first sill.
If the ground is good the crown-bars are placed entirely inside the brickwork of the tunnel, and they should not be so large as to equal half the space to be occupied by the brickwork. If the ground requires larger timbers than half the thickness of the brickwork, the bars must be worked as what are called ‘drawing-bars,’ that is, bars to be drawn on end as the brickwork progresses; or must be placed entirely above the brickwork of the tunnel and built in.
When the length is completed, and the invert taken out, two profiles of boards—that is, light frames representing the exact shape of the invert—are set at the right level and to the right lines by the engineers; the bricklayers then commence and build the invert of the tunnel. When that is completed, what are called side-frames are set to guide the bricklayers in laying the bricks for the side-walls of the tunnel up to the springing of the arch. When the side-walls of the tunnel are completed to springing, the ‘centres’ are set.
The centres adopted at the Severn Tunnel were wholly what are known as ‘skeleton centres;’ that Progress of the work—1881. is, centres made by bolting two or three thicknesses of elm planks together into an arch of 3 inches less radius than the interior of the brickwork of the tunnel. The centres would be set from 3 feet 6 inches to 4 feet apart, according to the weight they would have to carry; and the end ribs of each length should generally be thicker and stronger than the middle ribs, as they have to bear the weight of the crown-bars of the adjoining length. These end ribs are called ‘leading-ribs.’
When the centres have been set, one ‘lagging’ is placed on at each side. A lagging is an ordinary plank or batten, 6 or 7 inches wide, and 3 inches thick. Behind this the bricklayers lay over-handed the bricks of the arch of the tunnel, making good to the ground at the back of the arch as they come up; or if the ground is loose, to the polling-boards, which have to be left in. This system is continued on both sides at once, putting on one lagging at a time till the arch is completed, except about 18 inches in the crown. To complete this, which can only be done by one man, cross laggings are used, called ‘block laggings.’
These being very short, are only made of 1½-inch boards, resting upon the top laggings at each side; the bricklayer, placing one or two at a time, works himself backwards till at last he completes the length, or, at a junction between two lengths, comes into a small hole, just the size of his body, which is known as the ‘pigeon-hole.’ This pigeon-hole he .jpg)
has to fill up, and complete the arch brick by brick, working himself out of it.
When the arch is entirely bricked in a ‘break-up length,’ what is known as a ‘running length’ is commenced at one end of it. For this the top heading has been already driven, as before described, and then the bars are placed in the same manner as in the break-up length, propped with back props, and then with props upon sills; but, as one end of the bars rests upon the brickwork of the finished length, there is only one face to timber down. The face of each break-up and running length requires to be carefully timbered, and if the ground is loose, polled; and the sills require long struts to secure them against any pressure at the face itself.
When the running length is completed at one end of the break-up, another running length should be taken out at the other end, the miners working in one length, while the bricklayers are working in the other.
The advantage of the bottom heading is this: that all the mining done in every length is dropped down into the skips or waggons with but little expense in filling; and as the bottom heading can be pushed forward at the rate of from 12 to 20 yards per week, about every six or seven weeks fresh break-ups can be started at distances of 100 yards apart. Of course the distance between break-ups must be regulated by the speed it is desired to attain in completing the tunnel.
If the ground at the top of the heading or tunnel Progress of the work—1881. is loose, especially if it is full of water, a different system must be followed to secure the top.
Before mining to place a head-tree, or a crown-bar, polling-boards must be driven by mauls as piles; and where the ground is wet, this is one of the most difficult operations miners have to perform, for the dirty water streams over the end of the piles, and at every blow of the maul is spattered on all the men that are near.
A considerable length of the Severn Tunnel, on the Gloucestershire side, was in loose gravel, full of water, and required this operation.
In the same ground the crown-bars had to be placed entirely outside the tunnel, and the brick-work of the arch completed under them, the space between the crown-bars being filled up to the polling-boards with rough brickwork or rubble.
In shaft-sinking, I have known cases more than thirty years ago, where, when a depth, say 20 or 30 feet, had been sunk from the top, a curb was placed, carried by iron rods, and in some cases by chains from timbers laid across the top of the shaft, and the brickwork for lining the shaft was built upon this curb; and in some cases the brickwork has been built continuously on the top of the shaft, and the lining lowered away till it reached the required level. In other cases short lengths have been taken out in the sinking, and the brickwork added below the first curb on other curbs placed from time to time.
.jpg)
I have, however, adopted the system of completing the whole of the sinking of the shaft before commencing any of the brickwork. In doing this, strong curbs are made similar to the centres of the tunnel; those for an 18-ft. shaft being 21 feet in external diameter; those for a 15-ft. shaft being 18 feet in external diameter; and as we ultimately sunk a large pumping-shaft, 29 feet in diameter, with 3 feet of brickwork, the curbs in that case were 35 feet in external diameter.
These curbs are placed from 3 to 5 feet apart as the sinking of the shaft progresses, with polling-boards behind them to support the ground; and when the shaft is completed to the bottom the brickwork is commenced, and each curb taken out as the brickwork built up from the bottom reaches it.
In sinking a shaft at any place on the Severn Tunnel, the principal expense was incurred in keeping the shaft free from water. We have tried all kinds of pumps for this purpose, and there are objections and difficulties with all.
The pulsometer-pump, which can be slung upon chains and lowered as the work progresses, with a flexible rubber hose at the bottom of the pump, can only be used to a depth of about 50 feet, and even at that depth consumes so much steam as to be very expensive.
A direct-acting steam-pump, of which we had many patterns, made by a variety of makers, could also be lowered in the shaft as the work progressed. Progress of the work—1881. It is continually getting out of repair, on account of the quantity of grit and dirt to be raised with the water.
A chain-pump can be used only for low lifts, say of 30 or 40 feet. Even the ordinary bucket-pump, which is the only one that can be trusted for working to depths of 100 to 200 feet, is subject to be constantly damaged and broken by the blasting in the pit. It is necessary, also, to have the holes of the suction-piece or wind-bore of the bucket-pump close to the bottom of the pump, or the miners would be standing in a considerable depth of water, and the pumps rapidly wear out their buckets from the quantity of sand and grit raised with the water.
In four of the shafts we sank at the Severn Tunnel, after we had a first shaft down and pumps fixed in it, we escaped this pumping difficulty by driving a heading from the existing pumping-shaft to the shaft we were sinking, or rather to the spot where the shaft would eventually be when it should be sunk its full depth. From this heading we sunk the bottom part of the shaft and lined it with brickwork, and then put down a bore-hole through the upper strata till it reached the completed work below; the water then drained through this bore-hole, as the men sunk the shaft, into the heading beneath, and so into the other shaft alongside from which it was pumped. In two of the shafts, by driving this cross-heading before the bore-hole was made, and accurately setting out the centre of the shaft .jpg)
underground through this heading, we succeeded in putting in the bottom and 20 or 30 feet of the shafts without having any water to contend with.
In a heading being driven at any distance from the shaft, though there may be but five or six men in the heading, the heat becomes very great, and, of course, a constant supply of fresh air is required for the men themselves.
To supply this, air-compressing engines were worked upon the surface. Besides some smaller ones, which had been provided by the Great Western Railway Company before the contract was let to me, we had two double 20-inch air-compressors, and one double 16-inch air-compressor. One double 20-inch was worked at the main shaft at Sudbrook, one 20-inch at 5 miles 4 chains, and the 16-inch one at the Marsh Pit.
The air from the compressors was led to an air-receiver, a wrought-iron cylinder, something like a boiler, in which the air was maintained at a pressure of about 60 lbs. to the square inch. Under this pressure the air was so hot that it was hardly possible to bear the hand for any time upon the outside of the receiver. From the receiver wrought-iron steam-pipes, 2 or 3 inches in diameter, were led away to the various places where rock-drills or other machinery were to work.
As almost all the headings required blasting, the boring of the holes was done by machine-drills. Several patterns of these were used: principally Progress of the work—1881. drills made by the Great Western Railway Company at Swindon, from a pattern of their own. This proved to be an excellent drill for making rapid progress; but it was complicated in design, and required constant repairs.
Another drill used was the ‘Darlington,’ which we liked better, because it required so little repairs. One of these drills could work for six months, after being started, without being repaired at all.
In addition to these we had several others, among them the ‘Duncan’ drill, which did excellent work.
Several methods have been tried for holding the drills in working a heading.
I think decidedly the best is the system adopted by the Great Western Company originally, of a heavy casting, called, from its shape, a ‘harp,’ with sliding bars that can be placed at any angle, and each bar carrying drills. These harps were made to run upon a strong frame on flange wheels of the same gauge as the road used for the skip-trolleys.
They could be pushed forward into the face, and when the holes were drilled, and before they were charged and fired, could be run back a short distance, out of danger from the blasting, and brought forward again when the material brought down by the blasting had been cleared up.
We used all kinds of explosives in the work: dynamite, gelatine, tonite, gun-cotton, compressed powder, and ordinary powder.
Powder made so much smoke as to be very Progress of the work—1881. objectionable in the workings, and was not strong enough for the hard rock.
The fumes of dynamite were so deleterious, and even dangerous, that we abandoned the use of it altogether.
Gelatine produced the best results in the hard rock, but we only used it for some months, when the makers were required by the Board of Trade to suspend the sale of it, and return all the material to store.
Nearly all the work was done with tonite made by the Cotton Powder Company. This is a carefully prepared explosive, made up into handy packages which the miners know as ‘pills,’ with a detonator attached to some of the cartridges, which were then known as ‘primers.’
The fuse was attached to a primer, and then, according to the depth of the hole and the strength of the material to be blasted, a primer and one, two, three, or four pills were placed in the hole.
Very little tamping was required; but in very hard rock the tonite would leave what the miners call a ‘socket;’ that is, a short length of the hole at the bottom where the explosive had not been strong enough to dislodge the rock. When using gelatine we found no sockets left, showing the extra strength of that explosive.
The tonite was not injured by water, nor did it seem to be affected by cold, as dynamite is; and the fumes were so slight that it was quite possible to Progress of the work—1881. return to the face within a minute from the time the shots were fired.
When the works were about half completed, I was persuaded to try compressed lime-cartridges, with which it was stated good work was being done in the coal-mines.
A number of these cartridges were brought down and tried in various places, but with no favourable result.
Generally in ground so hard as that we had to deal with, the lime only blew out its own tamping, and displaced no rock. The experiment, too, was attended with an unfortunate accident. The cartridges of compressed lime are placed in a large bore-hole, drilled to receive them, and a tube, instead of a fuse, is placed in the hole above the cartridges, and when the tamping has been placed, water is forced by a force-pump through the tube on to the cartridges. The swelling of the lime and the generation of gas is to displace the rock.
When one of the holes had just been loaded, my principal foreman, Joseph Talbot, was standing opposite to the hole, and the first of the water had just been pumped in, when the tamping blew out and a quantity of lime was forced into his eyes, and for a day or two it was feared he would lose his sight. Fortunately he entirely recovered; but the short experience was quite enough to convince us that we could make no use of the lime-cartridge system.
Progress of the work—1881. When the machine-drills were being worked by compressed air in any of the headings, the air, after passing through the drills, kept up efficient ventilation.
We afterwards applied this compressed air to the pumping. When we were sinking the drainage- heading, 20 feet below the old workings, it was done from a number of small shafts sunk to a depth of about 25 feet from the old heading; and to pump the water from the lower level I adopted a pump which was designed by Mr. A. O. Schenk, and which proved to be thoroughly efficient. The pump consisted of a wrought-iron closed tank of any suitable dimensions, say about 3 feet square. A 6-inch wrought-iron pipe, passing from near the bottom through the top of the tank, was led up to the level at which the water was to be delivered. A 1½-inch wrought-iron pipe was connected from the compressed air-main to the top of the tank. At the side of the tank, near the bottom, several flap-valves were fixed, opening inwards.
The tank was then lowered into the bottom of the shaft, which was 4 or 5 feet below the level of the bottom heading, and the connection with the air- pipe was made. The water flowed into the tank through the flap-valves till the tank was full; the pressure of the air being then admitted at the top of the tank, it closed the flap-valves and forced the water up the 6-inch pipe to the required level; when the tank was empty the air-pressure was turned off, and the tank filled itself again.
Progress of the work—1881. The only thing necessary was to make this self-acting, and this was done by Mr. Schenk in a very ingenious manner. A small wooden box, divided into two compartments, was fixed to rock on a centre at the top of the shaft, and a small pipe led to discharge directly over its centre. When this small pipe filled the one compartment of the box, so that it was heavier than the other, the box turned upon its centre-pins, and in doing so opened a valve on the air-main by means of rods attached to it, and allowed the compressed air to pass into the tank and force the water out of it to the surface. The small pipe then filled the other compartment of the box, and reversed the action, so that the pump was made perfectly self-acting, and required no looking after. These pumps were used throughout the period we were lowering the heading, and proved a complete success; and afterwards, when we had rather more water at the Marsh Pit than three 15-inch pumps could master, I fixed one of them there to lift water to a height of 100 feet. It was as perfect a success, lifting water 100 feet as it had been lifting 25 feet.
Of course, working by compressed air in this manner is expensive; but underground, on account of the difficulty of ventilation and the heat, it is necessary in many places to adopt compressed air for working machinery. Electricity could be used for the same purposes, but I have only used the compressed air, and that I have used for winding by steam-crabs, for pumping, and working rock-drills.
Progress of the work—1881. When the heading is driven, and a number of break-ups commenced along it, the great difficulty will be to take out the requisite number of skips of rock or other material filled from the lengths and the heading-face, and to take in the empty skips to be loaded, and to carry to the bricklayers the bricks and cement which they require; while at the same time there will constantly be timbers to be taken in to the miners.
Where the length to be travelled was short, this presented no great difficulty; but under the river about a mile of the work was done eastward from the shaft at Sudbrook, and about 1¼ miles westward from the shaft at Sea-Wall.
For the length done from Sudbrook, the gradient falling towards the shaft, there was laid up the 9-ft. barrel, throughout the whole of its length, a double road of 1-ft. 9-in. gauge, so that the skips could travel on the up or down road; but beyond the end of the 9-ft barrel, where the break-ups were commenced, the road had to be arranged in a series of single roads and turn-outs or sidings, to allow the skips to pass each other. The roads being thus arranged, the hauling was done by stout ponies from 13 to 13½ hands high, and these ponies became most intelligent at their work, and knew exactly what to do, even as well as the men themselves.
On the other side of the river, however, where all the material had to be taken out up a gradient of 1 in 100, it was much more difficult to take out the Progress of the work—1881. excavated material and to keep up the supply of bricks, cement, and timber; and as soon as the first length of half a mile of tunnel was completed, I put down a hauling-engine at the top of the Sea-Wall Shaft, to take out the skips and bring in the bricks and cement. This hauling-engine had two cylinders, each 12 inches in diameter, and worked a large pulley, round which three turns of a wire-rope or bond were taken; the two ends after leaving the pulley descended the shaft, and there passed round other pulleys to change the direction. They then ran down the tunnel, first for half a mile, and eventually for a distance of rather more than a mile from the shaft, and at the extreme end passed round another pulley, and the roads on which the skips travelled were laid as far apart as the diameter of this large pulley. Steel rollers were fixed on the sleepers of these roads several feet apart, over which the ‘bond’ ran; and at the top of the shaft there was attached to the rope a heavy trolley, carrying a considerable weight of iron placed upon a sharp incline, the weight of this trolley on the incline serving always to keep the wire-rope sufficiently tight. When the engine was started there was an endless wire-rope constantly in motion at a speed of about 2 miles an hour; the one line of wire-rope running down the tunnel and the other line up. The skips were brought out by ponies or men to the end of the rope, and there attached by the ‘hookers-on’ by means of a clip to the wire-rope, without stopping it.
Progress of the work—1881. The descending skips were pushed off the cage at the bottom of the pit, and were in like manner attached to the descending rope by another man. The skips were thus attached to the rope at any irregular intervals just as they came, without loss of time; and sometimes there were as many as 100 skips on the ascending rope and 100 on the descending one.
We found the system to work perfectly, and it reduced the cost of hauling out to less than half the cost of ponies.