Sapperton Railway Tunnel

Sapperton tunnel
Gloucestershire
Railway tunnels with cutting
Location
Carrying:	Golden Valley Line railway
Location
Grid reference:	SO938023
Location:	51°43’12"N, 2°5’27"W
Structure
Length:	1 mile 104 yards
History
Information

The Sapperton Railway Tunnel in Gloucestershire carries the Golden Valley Line through the Cotswold escarpment near the village of Sapperton after which it is named. The railway line runs up the Golden Valley from Stroud, and is then faced with the precipitous side of the valley, through which the Sapperton Tunnel bores for over a mile to carry the line on towards Swindon in Wiltshire.

The tunnel was begun by the Cheltenham and Great Western Union railway in 1839 and taken over by the Great Western Railway in 1843, being completed in 1845.

There are actually two tunnels, with a brief gap where the line runs in a cutting:

• Main tunnel; 1 mile 104 yards: 51°43’25"N, 2°6’6"W to 51°42’60"N, 2°4’48"W
• Second tunnel, 353 yards: 51°42’59"N, 2°4’45"W to 51°42’54"N, 2°4’30"W

Construction, engineering, and maintenance difficulties

The initial plans for the tunnel, dating from 1835, were unusual in that it was proposed to construct the tunnel on a curve, but this seems to have been abandoned before any construction was done; some works remain which are thought to relate to the approach route for the original line, but no excavations were made on that line for the tunnel itself. In 1836 "Mr Brunel" was appointed as engineer for the project; this refers to Isambard Kingdom Brunel, but the involvement of his father, Marc Isambard Brunel, is also recorded. Brunel promised to get rid of the "objectionable" curve, and plans of the revised straight alignment were deposited in 1838. Preliminary shafts were dug, the work beginning in 1837, to ascertain the geological conditions, on the same straight alignment on which the tunnel was eventually built. In 1841 work began on four additional shafts of a larger diameter, plus a trial heading along the tunnel alignment. The line was opened in 1845.

Geological cross-section of Sapperton railway tunnel

It was found that the intended route passed through a layer of fuller's earth, which was not sufficiently stable to allow construction of a tunnel. The plans were therefore revised to situate the tunnel in more stable strata at a shallower depth, at the expense of steeper gradients on the approaches. This also reduced the length of the tunnel from 2,830 yards, on a 1 in 352 gradient in the 1836 plan, or 2,730 yards on 1 in 330 in the 1838 plan (1 in 90 in the main tunnel, as built, and 1 in 57 on the approach to it), and so reduced construction costs; a 1950 article, based on the original contract and specification, quoted the directors as saying it would, "considerably diminish its length and expense" and asserted that this was the actual reason for the change to the shallower depth, rather than simply an effect of it. The 1950 article said that the drainage through the oolite was so good that construction was probably started from the foot of each shaft without drainage. Brunel was quoted as reporting in October 1841 that, "the drainage of the water is obtained into the lower oolite, without pumping, at one of the intermediate shafts" Therefore, although it included the geological cross-section (drawn by GWR engineer, Mr. R. P. Brereton), it made no mention of fuller's earth problems. Later articles say the section shows that the header, on the originally-proposed deeper level, passes through a much greater length of fuller's earth than the tunnel as built, and that it also makes it clear that information gained from digging the shafts would have made it apparent that this would be the case.

This diagram also disproves the unsupported assertions which are sometimes made that the gap between the two tunnels is the result of a roof collapse in the early days of the tunnel. This gap coincides precisely with the level section at the summit of the line, and also with a dip in the contours of the ground above which brings the ground level below the depth at which the transition from tunnel to cutting is made at the outer ends of the tunnels. It is unlikely that a chance collapse would have taken place to correspond with these features with such convenient precision. The gap is also shown to be in the more stable oolitic strata rather than the unstable stretches of fuller's earth which are more liable to collapse.

The lateral alignment of the revised tunnel route was the same as the planned deeper route, and this left the ten exploratory shafts intersecting the tunnel and continuing as pits up to 6 metres deep below the tunnel floor. The pits were capped with timber to support the track and ballast. However, no records were left as to whether or not the pits had been filled in.

In 1950 a train driver noticed a void beneath the tracks and it became apparent that work was required to stabilise the pits. The original timbers were removed and the pits spanned with prefabricated concrete beams reinforced with bullhead rail and shear links. One more beam was fabricated than was required for the work and this was stored in a nearby yard. Again it was not recorded whether the pits were filled.

In November 2000, heavy flooding caused the collapse of one of the capped shafts. Emergency work over a period of four weeks was carried out to stabilise the collapsed shaft and similar preventative work was carried out on some of the other shafts; the four larger shafts, however, were left untouched. An erroneous report attributes the damage to the collapse of the canal tunnel running underneath: in reality the canal tunnel is to the north of the rail tunnel and does not pass underneath it.

In 2001 it was decided to test the redundant beam remaining from the 1950 operations to destruction in order to determine its strength, as part of surveys intended to discover whether the route would support the loadings required: it failed the tests and so the tunnel was closes and work carried out in a major operation over seven days by over 100 men working 12-hour shifts, many accommodated in temporary buildings at the site. The 1950 beams were cut up and dragged out and each pit was spanned with new prefabricated beams, joined using stainless steel bolts.

Not until the old beams were removed did it become known that the pits had indeed been backfilled in the course of earlier work and the fears of falling into a 20-foot hole were unfounded.

Accidents

A collision took place near Sapperton Tunnel (mis-spelt as "Salperton") on 4 December 1851. A goods train approaching the tunnel from the Swindon direction was overcome by the gradient after passing Tetbury Road station, and the driver decided to divide the train, taking the front portion forward and returning later to collect the rear portion. Unfortunately the brakes on the rear portion failed and it ran back down the gradient to collide with a following train. The goods vehicles were destroyed and the driver and passengers of the following train sustained injuries, but there were no fatalities.

A minor accident occurred at the tunnel on 29 October 1855 also involving a train becoming divided and one portion running away. There were no injuries or deaths and it is thought that no investigation was carried out.

Four platelayers were killed in the tunnel on 14 April 1896. This is mentioned in Hansard for 27 April 1896 (the report mis-spells "Stroud" as "Strood") but no details are given.

Disruption to train services was caused on 17 January 2011 when a huntsman and twenty fox-hounds trespassed on the line near the east end of the tunnel. One of the hounds was struck by a train and killed. The train was cancelled and other trains suffered delays. The trespassers had disappeared by the time British Transport Police officers arrived, and no hunt admitted responsibility.

See also

• Sapperton Canal Tunnel, which runs very close to the railway tunnel.

Outside links

("Wikimedia Commons" has material about Sapperton Railway Tunnel)