Atmospheric railway
BRUCE HEALEY charts the rise and fall of atmospheric railways… and ponders a possible rebirth
BRUCE HEALEY charts the rise and fall of atmospheric railways and ponders if they could be due a renaissance.
History has a habit of politicising and simplifying, and thus the general view of atmospheric railways is of a complete fiasco.
Yet two of the Victorian systems ran relatively trouble-free for over a decade, while the other two had decent periods of reliable operation. With present-day technology, a slow rebirth is under way. Atmospheric railways are the story of a ‘might have been’ and ‘might yet be’.
While the Rainhill trials of 1829 represented an important victory for steam, many problems still existed. Into the 1840s, accidents were too frequent, engines were unreliable, and there were difficulties with loaded trains on gradients - rope haulage was used on the entrance to Euston station, on the Blackwall Railway and the Canterbury and Whitstable, among many others.
Atmospheric systems appeared modern, able to cope with gradients, and safe.
The safety argument was based around there only being one train in each section at any one time. Steam trains, by contrast, were normally controlled by the time interval system, which was basically a man with a signal or flag and a watch. After a designated time, the next train was allowed to pass.
Shipbuilders Joseph and Jacob Samuda, briefly joined by Samuel Clegg, developed the system that would be generally adopted. A demonstration line with a gradient of 1-in-120 was built at Wormwood Scrubs in 1840. It ran for about two years.
An iron pipe laid between the rails had a narrow slot along the top, sealed by a leather flap with iron reinforcing plates. The first carriage carried a piston which, connected via a bracket, ran in the pipe. Wheels connected to the piston bracket forced the flap open and a wheel on the piston carriage closed it again. A mixture of tallow and beeswax was used to seal the flap against the pipe. Partial vacuum on one side of the piston and normal air pressure behind drew the train along.
Engine houses located near the end of a run of pipe had two engines which could run independently or be coupled together, which exhausted the pipe shortly before the train was due. The piston carriage featured a braking system.
Where a station was not located at the end of a pipe, the train was braked against the vacuum. This required considerable skill and overruns were frequent.
To start the train, the piston carriage needed to be moved to the end of the pipe, usually by
gravity. An entry valve had a train-operated slide valve which opened the vacuum to the end of the pipe. The valve was closed by the change of pressure.
Near the end of the pipe, the piston passed the connection to the engine house and the remaining air was compressed - opening an exit flap valve which was normally held shut by the vacuum. Because the pipe was mounted on the sleepers, the lowest part of the piston was below the level of the rail, making diamond crossings and points impossible.
The Dalkey Line
The first commercial atmospheric railway ran from Kingstown (Dunleary) to Dalkey.
The standard gauge Dublin to Kingstown line opened in 1834 and was an immediate success. A 1.75-mile wagonway was opened to Dalkey to transport granite for shipment, and the 1-in-110 gradient made it a suitable candidate for an atmospheric system, which opened in 1843. Robert Stephenson, William Cubitt, Isambard Kingdom Brunel and delegations from the continent were among the visitors.
There was one engine house at Dalkey. Train loads were up to 70 tons, including the piston carriage and passengers. An empty carriage would have weighed around five tons.
The train was manually pushed to the end of the pipe at Kingstown and held against the vacuum until ready to run. The pumping engines continued to run until a lookout spotted the train approaching Dalkey. The return trip was by gravity.
The line operated successfully until 1854 when a required change to 5ft 3in gauge and extension to Bray were imminent.
Many lines were subsequently proposed and in a rare moment of governmental clarity, a Select Committee was set up. Of the leading figures of the day, Cubitt, Charles Vignoles, and Brunel spoke in favour while Robert Stephenson, Joseph Locke and G.F. Bidder opposed.
The Board of Trade sensibly refused more schemes until the technology was proven. Croydon (with extension to Epsom) and South Devon were authorised.
The Croydon Line
With the Croydon Line carrying trains from the London and Croydon, the London and Brighton and the South Eastern Railways, capacity was being reached. The London and Croydon opted to build a third track for atmospheric trains, which opened in January 1846.
The atmospheric line was built on the east side of the existing lines. To reach West Croydon a wooden flyover was constructed over the London & Brighton lines, with a gradient of 1-in-50. As the eventual goal was to reach London Bridge, the alternative would have been a flyover over the Bricklayers Arms line on the London side of New Cross (later renamed New Cross Gate).
The flyover was a world’s first. There are a number of accounts of passengers having to get out because of a lack of vacuum, and sometimes push.
Initially, two sections of pipe were involved - a three-mile section from Forest Hill to
Jolly Sailor (an earlier station to the north of Norwood Junction station) and a two-mile section from there to West Croydon. There were three engine houses.
Speeds of 70mph were recorded with six carriages, and 30mph with 16. Public operation started at the end of February 1846.
The Croydon system had a bi-directional entry/exit valve. The request to evacuate the pipe was provided by telegraph.
At Forest Hill, where a change to locomotive power was necessary, a set of points was installed over which a piston carriage could travel, and which reportedly needed six people to operate it. The Forest Hill engine house had four stationary engines to speed up the change of motive power and as a reserve for the gradient up to New Cross.
At some stations, passengers from the steam trains had to cross the atmospheric line. A drawbridge plate was provided, which was operated by railway staff. Some narrow escapes were reported.
Trains were started by gravity. At intermediate stations between sections, the station had an apex allowing restarts in the same direction. At West Croydon and later at New Cross, it is believed that the piston was
dropped, allowing the piston carriage and the rest of the train to be shunted normally.
The original cast iron crankshafts of the Maudslay stationary engines all broke at one time or another. They were replaced by wrought iron ones.
The hot summer of 1846 led to flap valve leakage which caused the system to be stopped for six weeks from early June, in order to install an improved flap valve and a better tallow mixture. A drought in late summer led to some engine houses running short of water, causing either a lower vacuum and trains to run slower or a steam engine replacement service.
During 1846, the London and Croydon merged with the London and Brighton to form the London Brighton and South Coast Railway. The new board did not favour the atmospheric system.
By autumn, the service appeared to settle down. The Times in December reported that of 3,000 trains run over the previous three months, only 11 were late by 15 minutes or more. This did not take account of some services being locomotive hauled, because of problems.
The three-mile section from Forest Hill to New Cross, which involved a 1-in-100 incline and another engine house, was late being commissioned. Testing started in August 1846, with opening in February 1847.
Gravity was used on the return trip. The 15-inch pipe and stationary engines were now inadequate on this section for ever-increasing loads and plans to extend the system to London Bridge depended on conquering the gradient.
The extension to Epsom was being built for both locomotive and atmospheric operation, but by December 1846 the decision had been made to use steam trains only. In May 1847, the LB&SCR board had had enough, and atmospheric operation was abruptly stopped.
Latterly, 39 trains a day had been working with decent reliability. In the context of the 1840s, that was quite an achievement. The atmospheric system and flyover were dismantled with some haste.
The South Devon Line
Brunel suggested atmospheric working for the single-track South Devon Line from Exeter to Plymouth and on the branch to Torre, with work starting in 1844.
On the 20-mile stretch from Exeter to Newton (later renamed Newton Abbot), there were seven pipes with eight engine houses. Beyond Newton, work was started on the Totnes and Dainton engine houses. There were gradients as steep as 1-in-36 beyond Totnes.
Brunel specified 13, 15 and 22-inch pipes for the various sections of the line, according to gradient. The 13-inch pipe was replaced before operation started. Late delivery and commissioning caused locomotive haulage to be used initially.
Trains were started using an eight-inch pipe to one side of the operating lines with a hazardous rope connection to the train. At stations which were not at the end of a pipe, Brunel had built wooden platform extensions to cover overruns. He also designed level crossings with a vacuum-operated plate covering the pipe.
The first public atmospheric trains ran in September 1847 as far as Teignmouth. (By this time, the Croydon system had been abandoned.) Between November 1847 and the following March, services to Newton were built up. At times a reliable service ran.
The leather valve caused problems in freezing conditions. By summer, corrosion in the metal plates in the valve was leading to tearing of the leather. On all atmospheric
systems, the tallow needed regular reapplication and the valve needed frequent checking - all adding to the cost of operation.
The telegraph system to request the vacuum was delayed, meaning that the pumps were started according to the timetable rather than to the actual train times, incurring extra cost.
With the high running costs, increasing loads, the unreliability of the flap valve and the limitations of single track, the atmospheric system was abandoned in September 1848.
By this time, work on the next sections had progressed with the Dainton engine house (using engines originally purchased for the Epsom line) all but complete, although the Totnes engine house was late.
The Saint-Germain Line
The line from Paris-St-Lazare to St-Germainen-Laye had opened in 1837. But it had the disadvantage that it terminated at Le Pecq, somewhat short of St-Germain, requiring passengers to walk over a river bridge and up a steep hill.
Work started in 1845 on an atmospheric system, which opened in April 1847. It was just short of a mile in length with a gradient of 1-in-28. A 25-inch pipe was used with an engine house at St Germain. Trains descended by gravity. A further section back towards Paris was never turned over to atmospheric operation.
Operation was relatively trouble-free, although in February 1858 the brakes failed on a train descending from St-Germain. It collided with a locomotive, causing three deaths. Steam operation took over in 1860.
Other systems
The Post Office:
Thomas Rammell and George Medhurst had both proposed a system which used the train itself as a piston - a large diameter reversible fan used air pressure to push the train outwards and vacuum to draw it back.
Negotiations with the Post Office led to a 30in tube to carry mail from the Euston arrival platform to Eversholt Street. It opened in February 1863. Two of the cars are on display at the Post Office museum at Mount Pleasant.
Further developments led to dual 1.75mile tunnels between Euston station and Cheapside. The tube was about 4½ feet wide and four feet high. It first ran in October 1865.
However, it did not offer much of a cost saving and services ended in October 1874. The disused tunnels were largely forgotten until just before Christmas 1928, when a gas explosion at Holborn disrupted half a mile of roadway and killed a workman.
The Waterloo and Whitehall Pneumatic Railway:
Plans were afoot for the first tube railways in London. A broad-gauge test line was set up at Crystal Palace in 1864, based on Rammell’s system to carry passengers through a tunnel. The carriage entered the tunnel by gravity and air-tight doors closed behind it.
Based on this, the Waterloo and Whitehall Pneumatic Railway was authorised and work started in October 1865. A trench was to be dug across the Thames into which pipes would be lowered and lined. But with the work not long under way, a financial crisis brought the works to a permanent halt, leaving wooden piles jutting out of the Thames.
But the story of atmospheric railways didn’t end there.
Aeromovel: Towards the end of the 20th century, an automated system was developed by Aeromovel Corporation of Brazil. Rails are mounted on an elevated hollow square section concrete box girder, which forms the air duct. Electric air pumps use both pressure and vacuum to work the train.
The first implementation was in 1989 in Jakarta (Indonesia), to serve a theme park.
A second system opened at Porto Alegre Airport ( Brazil), and a further incomplete system for the Brazilian city of Canoas started construction in 2016.
Hyperloop: The conceptual use of vacuum ahead and pressure behind the pod is more about reducing air resistance than propulsion. Linear motors provide the main drive.
Conclusion
For a heavy rail system, atmospheric propulsion is unsuitable. The Dalkey and St Germain systems survived longer than others partly because they were short and 50% operated by gravity.
‘Suck and blow’ systems pack more power into the same cross section of pipe, but not enough for heavy trains. The complications of points and crossovers make an extensive atmospheric system unworkable.
However, for light rail systems, atmospheric power seems a more reasonable solution. Perhaps atmospheric railways are due for a renaissance?