Wonders of World Engineering

 © Wonders of World Engineering 2015  |  site map  |  info@wondersofworldengineering.com

The building of the tunnel beneath the River Severn is one of the epics of engineering. For nearly fourteen years men worked on an underwater short cut between England and South Wales, waging a perpetual battle against inundation

ENTRANCE TO THE SEVERN TUNNEL on the Gloucestershire side of the river

ENTRANCE TO THE SEVERN TUNNEL on the Gloucestershire side of the river. The tunnel has a length of more than four miles, of which about two and a quarter are beneath the estuary of the River Severn.

THE building of the Severn Tunnel is an epic, a record of hope and heartbreak, of fourteen years’ fighting against the forces of Nature deep down below the bed of Great Britain’s longest river. Years of careful planning preceded the building of the tunnel.

In the earlier days of main line railways, the widening estuary of the River Severn formed a large and serious rift between the systems of southern and western England and the rich industrial country of South Wales, and this rift could be crossed only by a steam ferry service. The ferry involved two tran-shipments for all land-borne goods and passengers entering or leaving South Wales on the southern side.

But either of the alternatives, to throw a bridge across or to drive a tunnel beneath the river, promised to be a difficult task, even ruling out unforeseen difficulties, which generally crop up all too plentifully in such undertakings. As far back as 1862, however, an engineer engaged in building piers for the ferry steamers looked at the rolling width of the Severn, and decided that a tunnel beneath its bed was possible. That engineer was Charles Richardson.

It was a tremendous work that Richardson visualized, additionally heavy in those days when boring tools were of the simplest kind, and when engineers lacked the experience gained in boring the long tunnels of later years. Richardson reckoned that the tunnel could follow much the same line as the existing ferry carrying the Bristol-South Wales traffic. The tunnel would be about four and a half miles long, the underwater section approximately two and a quarter miles long, that being the width of the river at this point. The considerable extra length would be necessitated by the gradual downward inclines to a level safely below the bottom of the deepest part of the Severn.

Having made a number of careful observations, Richardson put his great-project before the officers of the Bristol and South Wales Union Railway, which was later to become a part of the Great Western Railway system. The Union welcomed the plan, and in the following year, 1863, a Bill for the building of a tunnel under the Severn came before Parliament. It was difficult to induce capitalists to back the scheme, and twice the Bill failed on that account. Finally the rich and resourceful Great Western Railway took up the scheme in 1872. John Hawkshaw was called in as Consulting Engineer and for a third time the project came before Parliament. There was something of a struggle for the Bill but, backed by the clear and convincing evidence of Hawkshaw and by the imposing sponsorship of the railway company, the promoters obtained their Act.

The bed of the Severn at the critical point rests on firm and closely packed strata but, as might be expected, it is not entirely even. At high water the river surface stretches from bank to bank, but at low water the irregularities of the bed become visible. The main stream passes over and through a channel called the Shoots, which it has carved out of the underlying strata in the course of thousands of years. At low water there are large stretches of bare land on either side of the Shoots, the channel of which lies nearer the Welsh side.

The county of Monmouth, though largely Welsh in character, is strictly in England, but the Gloucestershire and Monmouthshire portals of the Severn Tunnel are always referred to as English and Welsh respectively. To build a safe tunnel the engineers had to allow for a considerable inclination to clear the bottom of the Shoots.

Charles Richardson arranged for the crown of the tunnel arch to be situated 30 feet below the bottom of the Shoots, the tunnel rising from this maximum deep level on an inclination of 1 in 100 to either portal. The Severn has a tendency to flood because of the opposition of a heavy flow coming down from the Welsh hills with the famous bore, the high tidal wave that sweeps up the channel from its mouth.

The first requisite was to guard against the possible flooding of future tunnel workings by an overflow of the river. To combat this, which threatened on the low-lying English side, Richardson and his men threw up great dikes and embankments to retain the river in unusually high tides.

On March 18, 1873, Richardson made a start on the tunnel itself, beginning at Sudbrook, on the Welsh side. A 200-feet shaft was sunk down through the marl and shale into the underlying stratum of limestone. The diameter of the shaft was 15 feet. From the bottom of it was driven a heading, or pilot tunnel, ultimately intended for drainage purposes, eastwards under the river. This heading was 7 feet wide and rose from the bottom of the Sudbrook shaft at an inclination of 1 in 500.

It was no easy task, boring through the hard strata of limestone and shale with tools such as would be considered crude and antiquated to-day. Little by little they penetrated farther below the river, making for a point below the bottom of the Shoots. Their progress was slow.

Not until four and a half years had elapsed did they reach a point beyond the bottom of the Shoots, having by then penetrated for a distance of 1,600 yards from the bottom of the Sudbrook shaft.

“Talking” Timber

Again and again the workmen tapped underground springs. Imagine them, plying their picks in the narrow space of the heading, by the dim light of guttering candles, and grotesquely clad in flannel “donkey” jackets. Above and round them water oozes, trickles and drips from the surrounding rock. Under the huge pressure of the overlying strata, the timber props of the heading are constantly “talking” - emitting ominous groans and cracklings fit to terrify the uninitiated visitor to the workings.

In normal conditions the 1 in 500 inclination of the heading was sufficient for it to drain itself without undue inconvenience to the men working in it, but the flow of the percolations, as it reached the bottom of the shaft, had to be removed by constant pumping, carried out by steam engines on the surface above. After the sinking of the first shaft, Richardson sank others, one through the gravel and marl on the English side, close to the embankment holding back the high water, and the others through the alluvial beds on the Welsh side west of the Sudbrook shaft, or Old Shaft, as it came to be called. From these new shafts the tunnellers excavated further headings along the line of the future tunnel. Year after year they toiled far down below the level of the river and its surroundings.

By October 1879 it looked as if little time would elapse before the English and Welsh headings met below the bed of the river. On October 18 they had penetrated to within 138 yards of each other. But on that day the workmen in the heading from the Old Shaft on the Welsh side tapped a great underground spring. Thousands of gallons gushed out over the workers, extinguishing lights and bowling over skips and barrows. There was nothing for the men to do but run, and they bolted for their lives down the long heading to the bottom of the shaft. Had they hesitated they would have been drowned. They had no time to wait and close the great iron door in the heading which would have confined the water to its upper end. The heading filled up; the shaft filled up. The water rose through the shaft until it reached the tidal level of the Severn.

Sectional diagram of the Severn Tunnel

SECTIONAL DIAGRAM, drawn to scale, showing the position of the headings when the Big Spring broke in on October 18, 1879. The tunnel was planned for a total length of 4 miles 624 yards.

Sectional diagram of the Severn Tunnel

THE BIG SPRING BROKE IN at a point not far from the Sudbrook shaft and flooded the workings. Despite the great danger, Diver Lambert descended the shaft and after many attempts succeeded in closing the iron gate which would have confined the flood waters to the upper end of the heading.

Under-river working on the Welsh side was completely at a standstill. The formidable Big Spring, as it came to be known, cont-inued to discharge water into the workings at a rate far faster than the existing pumps could pump it out. It was in the face of this catastrophe that one of the great epics of engineering history took place. Nothing could be done with the flooded workings until that iron gate was shut, and the flood could be confined to the uppermost end of the heading. Nobody but a diver could hope to reach the water gate, and that diver needed to be a thoroughly experienced and superlatively courageous man.

In those days diving appliances were far less suited to deep-water work than they are to-day, and such a mission as the flooded Severn Tunnel workings would involve a heavy strain on the toughest constitution. The contractors found their man in Diver Lambert, a man of quiet habits, few words and tremendous grit. Lambert said he thought he could shut the water gate. Hoping and fearing, they lowered him down the flooded shaft, and he vanished into the dark depths until only the trail of silver air bubbles remained to show that he lived and breathed far below the surface. From the bottom of the Old Shaft, Lambert went cautiously along the pitchy dark length of the heading, with the tremendous pressure of the water surrounding him. He had to pick his way over deserted tools and barrows, moving slowly forward, step by step.

It was his air pipe that defeated him. It needed the greatest care and skill to avoid severing the frail tube against some sharp edge. One breakage or stoppage of the pipe and Lambert could never hope to see daylight again. For three hundred yards the brave diver groped his way along the point he could not drag his unwieldy air pipe. The man was ready to succeed, but the apparatus failed. He had to make his way back to the shaft, there to be hoisted to the surface with his mission unaccomplished.

At that time an inventor called Fleuss was demonstrating his new diving dress and apparatus in London. The great feature of the Fleuss appliance was that it eliminated the necessity for trailing an air pipe, as a tank of oxygen was carried on the back of the diver in the same way as a pack. The contractors got into touch with Fleuss and induced him to attempt the task at which Lambert, with his ordinary diving dress and air pipe, had failed.

Diver Lambert’s Heroism

In due course Fleuss arrived in Monmouthshire and was lowered into the flooded workings. The self-contained nature of his apparatus enabled him to move in the heading without fear of cutting off his own air supply. But after one experience of the dark heading, with the great pressure of water all round him and all manner of unseen objects impeding his progress, nothing would induce Fleuss to pursue the quest further and he gave up without having closed the gate.

Now Lambert volunteered to make a second attempt, and suggested that he should borrow the diving dress and apparatus of Fleuss. At first Fleuss demurred, but after some persuasion he allowed Lambert to put on the dress, with its helmet and reservoir. Lambert at first made some trial dives to accustom himself to the use of the Fleuss apparatus before risking himself in the flooded deeps of the workings. Finally he considered himself proficient in the use of the Fleuss apparatus.

Carrying one’s own supply of oxygen is not so simple a process as might be thought. The dangers of having too small a flow from the reservoir are great, and too strong a flow causes a condition resembling intoxication. Down went Lambert on his final mission along the flooded heading once again, still having to grope his way over the debris of abandoned equipment, but no longer haunted by the fear of a severed connexion. After several attempts he succeeded in securing the gate against the flow of the Big Spring.

After this first big halt in the proceedings, Sir Daniel Gooch, Chairman of the Great Western Railway and an engineer of great experience himself, appointed as Chief Engineer in charge of the works Mr. - afterwards Sir - John Hawkshaw, whose evidence had been partly instrumental in getting the Severn Tunnel Bill passed through Parliament. Hawkshaw knew that the existing pumping machinery would never prove sufficient to deal with the flow from the Big Spring. New pumps, with a 38-inch delivery pipe, were bought, and an attempt was made in the summer of 1880 to lower one down the shaft. Unfortunately passage was impeded by parts of the existing pumping plant near the shaft bottom. Those in charge decided that the best plan was to start the pump working where it was, and in this way the flooded workings were to a great extent exhausted of water. Then an accident happened to the pump and the water rose again. Seven months’ work was undone in a short time. During this period the inval-uable services of Diver Lambert were constantly in demand.

BEAM ENGINES operate six pumps in No. 1 Pumping House at Sudbrook

BEAM ENGINES operate six pumps in No. 1 Pumping House at Sudbrook. These pumps deal with the waters of the Big Spring. The engines have cylinders with a diameter of 70-in and a stroke of 10 feet.

At the eastern end, too, water broke in through the roof of the heading. The water was found to be salty, showing that it came from the estuary itself. The leakage was traced to the tidal area known as the Salmon Pool. The hole was filled up with alternate layers of loose and bagged clay.

While the struggle with the water was going on, Hawkshaw decided to make a distinct modification in the original plans of the tunnel. The incline from the English side to the lowest level was left at 1 in 100, but the Welsh incline, passing under the Shoots, was re-planned on a gradient of 1 in 90, to afford additional security against the inflow of water. By this rearrangement the section of the tunnel lying under the Shoots was lowered by 15 feet, so that the crown of the tunnel arch would come 45 feet below the bottom of the river. Little by little, by dint of continuous pumping, the drowned workings were dried out. In 1882, nine years after the preliminaries had been started, men were put to work on the approach cuttings at either end of the coming tunnel. The flow of the Big Spring was partly dammed up and partly dealt with by pumping. The broken pump had been repaired late in the autumn of 1880. For a while the Chief Engineer thought he had the final word. Early in 1882 he made a tour of all the workings and reported satisfactory progress throughout.

Barrier Between Life and Death

By 1883 work was in progress on the whole of the tunnel, with the exception of a section 300 yards long in the neighbourhood of the Big Spring. The length of the tunnel had now been finally fixed at 4 miles 621 yards, of which approximately 2¼ miles were under the river estuary. The total length of the new works, from Pilning in Gloucestershire to Severn Tunnel Junction in Mon-mouthshire, came to 7 miles 748 yards. Then the Big Spring burst through its barriers again and surged through the underground workings at a rate of nearly 30,000 gallons a minute. The men had to swim for their lives, and some of them were drowned.

There was no alternative but to provide additional facilities for drainage and pumping. For this, Hawkshaw began to drive a side heading from the foot of the Old Shaft at Sudbrook, running parallel to the tunnel itself, to relieve the flooding. Meanwhile, he had further pumping appliances installed, by which it was possible to keep the main tunnel workings sufficiently clear for progress to be made through the remaining 300 yards of ground below the Shoots. Yard by yard the borers progressed, dripping wet with the continual seepage and flow from the only half-conquered Big Spring, which massed ominously behind its frail barrier. On October 12, 1884, when Sir Daniel Gooch himself was in the workings, the last section of the tunnel was driven through the remaining piece of rock. The Severn had been tunnelled from end to end.

A little over a month later it looked as if the Big Spring had been finally conquered, for on December 19 the new side heading penetrated the spring itself. This heading immediately took the flow from the Big Spring, leaving the workings of the tunnel dry. With the boring effected, though not completed, a start was made with laying down the railway tracks along the approaches, the English side being the first to be equipped. Progress was also being made with the brick lining of the tunnel, and this was com-pleted on April 18, 1885. With this apparent completion of the tunnel, Hawkshaw felt that he could safely dam up the Big Spring, now isolated from the tunnel proper and confined to the side heading. This he did by stopping up the side heading altogether and thus allowing the water to take its original course through the rock strata, round the outside of the tunnel lining.

On September 5, 1885, the first train passed through the tunnel from Severn Tunnel Junction to the English side and back. The distinguished guests included Sir Daniel Gooch. Unknown to those who participated in this interesting double trip, it was an exceedingly risky proceeding. Shortly afterwards the waters forced their way through the lining of the tunnel, blowing the new bricks in all directions and striking the opposite wall with tremendous impact.

ENORMOUS VENTILATING FAN which supplies air to the Severn TunnelFaced with this new calamity which, had it come a little earlier, might have been disastrous, Hawkshaw decided that it was futile to hope that he could permanently hold back the water from the borings. The only thing possible was to carry off the entire flow of the Big Spring with powerful pumps.

By the time he had completed these remedies, his pumps were capable of dealing with 38,000,000 gallons of water a day. Most of this was ejected into the Severn, though a small proportion was drawn off as feed-water for the pumping engines.

ENORMOUS VENTILATING FAN which supplies air to the Severn Tunnel. Installed in 1924, the fan has a capacity of 800,000 cubic feet a minute. The fan is 9 feet wide and has a diameter of 27 feet.

On November 22, 1886, nearly three months after goods trains had begun to travel through the tunnel, it was inspected by the Board of Trade and passed for passenger traffic. The pumping appliances were still not complete, though they were capable, as they stood, of dealing with the maximum known flows of all the springs encountered in the course of the tunnel. The completion of the pumping engines would bring their maximum capacity up to 66,000,000 gallons in twenty-four hours, leaving a wide margin for contingencies.

Shortly after the Board of Trade inspector had gone over the works, on December 1, 1886, passenger trains began to run over the whole length of the new line from Pilning to Severn Tunnel Junction.

Approaching the tunnel from Bristol, the line is carried down in a deep cutting through successive strata of clay, peat and gravel, through which lie the courses of ancient river beds. Where the marl succeeds the gravel beds, the tunnel begins, running at a long slant down through the marl into a thin vein of grey sandstone, followed by a thicker vein of red sandstone. Through the deepest section of the tunnel the course lies through various rock strata of different thicknesses, chiefly clay shale and blue shale, with a small coal measure for some distance on the English side of the lowest level.

Below the tunnel, and running inwards from the deepest point on the main Sudbrook shaft, runs a drainage culvert 5 feet in diameter and 1,166 yards long. This culvert ends under the eastern end of the 264-yards level section in the middle of the tunnel, whence the 1 in 100 and 1 in 90 sections rise to the English and Welsh sides respectively. Above this great drain, also giving off the Sudbrook shaft and immediately below the rising floor of the tunnel, is a heading for ventilation purposes, 9 feet in diameter and 286 yards long, fed from huge fans in the Sudbrook pumping station above.

The tunnel itself contains approximately 77,000,000 bricks in its lining, which is 27-in thick, and 37,000 tons of cement. On the downward slope from the Bristol direction, it passes first under a tidal lagoon known as the Salmon Pool, then under the reef called English Stones before reaching its intermediate level section under the Shoots. Beyond the Shoots it rises below an outcrop known as the Gruggy, divided from the high bank of the Monmouthshire shore by another tidal pool. Continuing up the western incline it passes under the little River Nedern, one of the lower tributaries of the Severn, before emerging in Monmouthshire.

[From part 20 and part 21, published 13 and 20 July 1937]

You can read more on the Severn Tunnel in Railway Wonders of the World (1935), and within an article on

the Great Western Railway from Frederick Talbot’s Railway Wonders of the World (1913).

Conquest of the Severn