THE ELEMENTS OF DANGER in RAILWAY CONSTRUCTION (Railways of Australia Network 1989) "The Natural Elements pose a Constant Threat to World Railways" The Ancient Greeks believed the world was made up of four elements earth - air, fire, and water. And they accepted that sometimes these elements were hostile. Over some 165 years, all four elements have thrown their worst at the railways, causing innumerable accidents, some so serious as to be woven into regional folklore some avoidable, some less so. Purists will say that no accident to a train caused by slides, gales, bushfire or flood is truly unavoidable. Near-perfection in design, maintenance, patrolling, communications and hazard detection equipment can and have prevented accidents in this "elemental" category. For example, the Shinkansen lines in Japan are laid in a country be,set by typhoons and earthquakes - yet in more than 20 years and millions of train movements they have never lost a train. The pessimist might add "not yet", but the Australian railway engineer would confirm that Shinkansen safety approaches perfection. But not all railways can justify the elaborate protection arrangements of these exceptionally busy and, compared with complete rail networks, relatively short superspeed lines. And there is a practical limit to what can be done. You can detect rockfalls with wire loops and floods with flood gauges, linking both into a failsafe signal system that will usually detect rupture of the track by, for example, a flood dropping a bridge into the river. But how do you detect the embankment or culvert that collapses right under the train? Or the culvert that washes out, leaving the rails in mid-air? Or the bridge that is hit by flood debris and is merely knocked out of alignment enough to derall a train? How do you detect the bushfire-weakened timber trestle or the freak wind gust that hits a three-tier motor car carrier carrier breadside-on? The answer, of course, is that you will detect only some of those "elemental" threats and then only with great difficulty. Usually it will be done less by gadget than as a result of intervention by an alert,dedicated human ,sensitive to danger and able to report it in time. The Sharp Lookout insisted upon by the Rule Book and the driver's valve of the automatic brake have saved far more trains than the fanciest natural disaster detection equipment. ENGINEERING. It is often forgotten that the earthworks on a road or railway are exactly that - works. They are engineering structures, disturbances by man to nature. Nature's original surface is itself sometimes unstable and often unable to bear the load of a heavy train without substantial preparation. Today this modification is known as the engineering science of soil mechanics, and- the construction of earthworks is a much more precise business than the apparent muddle of bulldozers and scrapers on a muddy construction site suggests. The geology - its bedrock, soils, slip planes and drainage - is studied carefuly. Cuttings (cuts) will be sloped, benched, and drained to take this into account. Embankments (fills) should be carefully built up from ,selected layers of material, often premixed. Each will be placed, rolled and compacted. Watering is a normal part of this. Special geotextile fabrics may be built into the structure between layers, or overlaid and sprayed with binder emulsions to achieve the right stability or to curb erosion. We have learned much from the highway engineers. Some of our railway engineers have even learned to leave tidy linesides that match the highway engineer's environmental sensitivity. In the bad old days, earthworks were never so good. Cuttings Were driven through bedrock, leaving soft soil overlaid on slope planes running towards the track, ready to slide with the first really hard rain. New lines were operated under freight traffic alone, so that their embankments could settle before passenger traffic was allowed at express speed. When they slipped, cutting works were dug out or shored with brick retaining walls; embankment slides were corrected by tipping yet more soft uncompacted fill. Gentle sinkages were made good by dumping ballast and lifting the track, which, given the cost of ballast, has always been an expensive solution; Failures of these unscientific earthworks soon occurred and, when undetected by humans, they caused serious moving-train accidents. Perhaps the earliest recorded land slip accident was in the much photographed Sonning Cutting, near Reading on the first section of Isambard Kingdom Brunel's 2130mm (7ft) gauge Great Western Railway. Around 5.30am on the frozen Christmas Eve of 1841, a down mixed train crashed into an earth slip. Eight people were killed and 17 hurt, most of them poor frozen wretches who had been huddled on the plank seats of the open low-sided .wagons that were the lowest class of travel in those hard days. The Board of Trade inspectors investigated under Mr Gladstone's first Regulation of Railway Act of the previous year, noted the gentle 2'.1 slope of the cutting, and reported in a manner kind to the GWR and the great Isambard. WORST SLIP. England's worst earthslip disaster was on the Midland Railway's Settle and Carlisle line. It had been surveyed over the bleak moors and falls by a young Tasmanian called Charles Sharland. The job broke his health, for he died at age 26 before the line was opened in 1876 as the last of the three great Anglo-Scottish main lines. On the winter afternoon of 19 January 1916, in the Long Meg cutting near Little Salkeld a snap thaw caused a landslip less than five minutes after the line had been patrolled. The Midland Scottish Express from London to Glasgow was pounding down the I in 132 gradient at 110km/h. It crashed into the slip, tipping 440 compound engine 1010 on to its side, and telescoping two of the gaslit wooden care into the tender. Fortunately, there was no fire (unlike two previous Midland wrecks on this line), but seven people died. The New Zealand Railways were less lucky in the early morning of 6 July 1920, near Ongarue. The Auckland to Wellington Limited Express, hauled by an engine with a less than brilliant acetylene headlight, ran into a large cutting slip of papa,(Papa is a very young sedimentary soft rock. Generally quite firm in situ, but breaks down readily to mud at uncovered surfaces when exposed to wetting / drying cycles. Often unstable in cuttings where lack of lateral support, exposure to wetting/drying, tree roots etc can lead to slipping along shear planes. The slip contained a large boulder. The engine went over and three gaslit wooden car telescoped. Gas ignited and the car burned until a second papa slip put the fire out. Seventeen passengers died and the Ongarue disaster led to a general conversion of all NZR locomotives and coaches to electric light. ' The Napier Express derailment of 22 September 1925 burned five more wrecked gaslit care, and kicked the electric lighting programme along, to the relief of politicians and the public and the profit of the American Pyle National Company and J. Stone and Co, who supplied the Iocomotive and coach lighting equipment. ROCKFISH. A more typica and recent landslide accident involved a modern all-stainless- steel train, the successor to the Southern Rallway's famous Washington-New Odeans Southern Crescent. At 5.20am on 13 April 1983, Amtrak Train 820, The Crescent, had rounded a curve in a cutting near Reckfish, Virginia, at 77km/h when the driver saw a shadow 70m in front of the train. The two GM diesels hit a landslip, derailed, jackknifed and piled up four of the six care behind them. As usual, modern couplers and strong structures saved the day. Nobody was killed. But 24 of 331 people aboard were hurt, and there was $US232 000 worth of damage. The cutting had been made for single track in 1860, to a 1:15 wall slope through a phyllite rock. In 1919, it was widened for a second track, which was lifted when CTC was put in in 1960. The slip was not large, only 76 cubic metms of sandy loam and silt,with 18 per cent water. The maximum depth over rails was well below coupler height -just 600mm. The cut had no record of instability, and it had been patrolled only six hours earlier. An explicable event? Or was it? The vegetation had been weakend by burning off, and there was ample warning of an instability problem from trees tilted ominously towards the track. The accident was at milepost 135.2. Just 14 months before,just one kilometre away at milepost 1359, the same Amtrak 820 had derailed on a mudslide in a cutting with a similar geology, slope, and history of natural vegetation clearance by burning. Rockfish illustrated how quickly a train (or at least its locomotive) will stop when it runs into an even modest earthslip. But how much soft earth and mud does it take to derail a train without actually stopping it? The answer,once again, is a lot less than most railwaymen believe, which highlights the vital importance to safety of keeping flangeways clear and unobstructed. At a rural level crossing near Hackettslown,New Jersey, heavy rain on 16 June 1925 washed some of the road's pavement formation over the Lackwanna Railroad's track. The material was soft mud, its depth only a few centimatres. This was insufficient to derail the driving wheels of a passing Pacific Locomotive. But the leading truck wheels were derailed, and 50 metms beyond the road crossing there was a turnout to a disused siding. The engine came off at the frog (crossing) of this tumout and ran down the bank, dragging two coaches and two Pullman sleepers of a spedal boat train excursion service beside it and on to it. It was then that the big Pacific's boiler ruptured. This turned a bad shake-up into Horror disaster in which 50 people died. The boat was the Republic and the White Star Line thoughtfully held her Sailing for the survivors to make their connection. WHAT TO DO? Control of the landslip hazard is a difficult problem for the distrlct angineer who has inherited kilometres of unstable earthworks, badly built decades before. What can be done? first a careful,case-by-case analysis of the local geology,soil and slip planes. Local drainage works,slope easing and benching can-help if there is room. Cement injection can help; the writer recalls it being done 40 years ago on a soft embankment in Sydney that still gives trouble. In some cases, the only solution is to build a concrete crib or a stout retaining wall. All of this is very expensive work, earning not a cent of extra revenue for the railway. But when passengers or dangerous goods are carried, it is work that is the unavoidable price of safety. The landslip accident reports, of course,cover the slips that caught a train because they wore net detected in time. For everyone of these accidents, there are hundreds of slip incidents, minor and major, that were not allowed to grow into accidents. These were caught by observant "eyes-up" patrolling of the cuttings, and by people who climbed off their trolley to inspect the banks and culverts below the line. Accidents were avoided by people armed with local knowledge, the training to look for the indicators of danger, common sense and a dedication to safety. There are few outdoor jobs on the railway that are less exciting and less glamorous than checking weed-covered earthworks that have probably been there for a hundred years. And few jobs are more unpleasant in wet weather. Yet few jobs are more necesarry for the safety of the line. ROCKFALLS. The walls of a cut through hard rock will usually be stable, even though exposed rock will weather and break up from the effects of wind, rain and ice. The dsks from small pieces of rock dislodging and falling on to the track are small and often they shatter on hitting the ballast or rails. The chances of them falling on a train are small and of serious damage smaller still. The real problems arise in places with unstable geology or large areas of rock that weather rapidly. Where the size and frequency of rockfalls cause anxiety, it is usual to protect the line with an automatic signal of come kind. A fence-like installation is common, with a series of light horizontal wires, typically 200-300 mm apart set on stout posts about 2m high, beside or above the track. The wires am connected in series. In the ealy days, the connection was mechanical via return pulleys; today it is electrical. Both operate so that a falling rock breaks at least one wire and sets the signal to danger. This works well, provided that the fall does not occur while the train is actually passing. No Australian installations to detect rockfall are known, but visitors to Scotland can see one in the Pass of Brander on the former Daledonian Railway's Callandar and Oban line, here the track follows the slope of Cruachan. This installation has generally given its warning in time - except in August 1946 when a large boulder fell on the track just in front of the 6.05am Oban-Glasgow express. With vacuum brakes applied in emergency, the engine crashed into it; lacking a cowcatcher, the locomotive lifted over the boulder and derailed itself, the tender and several coaches. Fortunately, nobody was hurt, for all-steel cars and strong couplers were not the norm in Britain at the time. Where major falls are a serious threat to the safety of the line, substantial protective rockfall shelters must be built over the track. Built in timber or steel, with an earth over- layer several metres thick, and today in reinforced concrete, these shelters function in exactly the same way as Swiss avalanche shelters on mountainsides. Rockfall shelters over track and tunnel entrances are rare in Australia, but examples can be seen on the steep 1 in 30 descent from Robertson to Unanderra on the Illawarra escarpment railway in New South Wales. MOVEMENTS. The heaviest earth movement is of course the earthquake, many of which have severely damaged railways. While no railway disaster due to earthquake is on record, railways in danger zones have been built, and alarmed, to cope with severe earthquake shocks. In Japan the Shinkansen high speed lines and the major subway stations in Tokyo are earthquake resistant. So, it is said, is the Seikan Tunnel,at 53kin long and 200m below sea level,once the world's longest and deepest; it cuts across an earthquake fault. So does BART, the Bay Area Rapid Transit system in San Francisco,where the Market Street subway runs right along the San Andreas fault - the fault that is predicted to again devastate San Francisco. BART's Transbay tube is also earthquake-resistant. In New Zealand, the new Rangitiki River bridges on the diverted North Island Main Trunk have their piers on and span a fault line. Of major Australian cities, only Adelaide is said to have any significant earthquake risk. In the unlikely event of this happening, repair and resupply roles for the railways have been built into the Australian government's national disaster planning. SUBSIDENCE. Mine subsidence is a railway problem quite close to home. But we have not seen anything remotely like the events on the Furness Railway at Lindal in England on 22 September 1892. Locomotive No 115, a Sharp Stewart 0-6-0, was minding its own business shunting wagons on coaling sidings when, without warning, subsidence created a funnel-shaped hole. No 115 rolled in, breaking fishplate bolts and rails. Fortunately, the engine sank without its crew. The railway company saved its tender, but FR No 115 went down hundreds of feet throught the loose earth and coal waste to become a total loss. It is still there, History does not record who bore the costs. In New South Wales, the Illawarra escarpment is honeycombed with old coal workings dating back to 1857. Many galleries were neither propedy surveyed nor the mining work documented. Such conditions apparently exist in the Stanwell Park area,where an elegant and much photographed eight-arch brick viaduct 148m long and 45m high sat on a 240m radius curve was built in 1919 to deviate the Illawarra main line, for 65 years the viaduct stood firm in its sylvan setting, until cracking was noted in the floor of arch 3 on 14 August 1965. It was repaired by strengthening the area of the failure, and the structure was very carefully watched by State Rairs civil engineers. On 8 December the up track was closed due to "loss of brickwork in the arch of another span", an SRA euphemism meaning that the span was starting to fall down. On 12 December 1985, just days before the opening of the Illawarra Line, both electrification,viaduct and line were closed. The remarks of the then chief executive, David Hill, and NSW Premier Neville Wran on being told of this most unfortunate timing are nat recorded. The challenge facing the SRA was to identify and fix an unidentified structural problem on a vital coal-expert and Sydney commuter railway with no plans, materials or information available, and with manufacturers and contractors shutting down for the Christmas holidays. The job was done with an exceptional level of dedication and engineering skill by constructing a piled trestle falsework under the faulty arch and dismantling it while under compression - the first known exercise of its kind. The failed span was replaced with a new steel and concrete deck slab structure,the repaired bridge was opened on 4 February 1986, but the structure has moved again since then. It is being carefully monitored, and major rebuilding or replacement will be needed. HAWKESBURY. Another form of movement underground forced the urgent and premature replacement of the original Hawkesbury River Bridge,the longest in NSW,during World War Two. The original double track structure, built by Union Bridge of New York and opened on 1 May 1989, was the third largest in the world at the time. Its piers, then the world's deepest, went down to 49m below water level, to what was believed to be solid rock. But in 1939, on the eve of the war, routine surveys found that one of the piers was tilting out of the vertical. It had been founded not on the rock underbed of the river but on a gigantic, boulder-like rock resting in a pocket of that bed. The bridge could not be closed, for there was a war on, and the Sydney-Newcastle line was not only the busiest intercity route in NSW, but a vital strategic link. It could not be repaired, for that too would involve closure to build two new piers and a long span to replace the two spans on either side of the defective pier and its "floating". foundation. So the NSWGR had to build a completely new bridge - and to do it in wartime, at the worst possible time for materials and skilled labour. The job was handled by the NSW Railways themselves and opened in eady 1946. Throughout the war, the two tracks on the old span were singled on the "high" side of the tilt and the trains kept to a 7km/h crawl over an increasingly defective bridge. With the old Hawkesbury Bridge, the NSWGR was very careful, and perhaps a bit lucky, not to create the wodd's worst rail accident caused by earth movement. That was to occur not in the open air or on a bridge but in a tunnel. On 17 June 1972, near Vierzy on the Paris Nord-Soissons outer suburban line,subterranean earth and rock movement brought down the roof lining of a tunnel. Into this crashed in rapid succession not one but two speeding diesel passenger trains. The impacts brought more rubble down on top of the wreckage, extending the area of the collapse. After five days' digging by volunteers, who worked under appalling conditions with further falls, the final count was made: 107 people lost dead. Dangerous stuff, earth, an element to be kept in its place and watched very carefully - especially when it rains. WIND HAZARDS. Winds above 30km/h will exert increasing severe lateral pressures on exposed surfaces, creating high sideways loads on bridge girders and toppling moments on piers. Similar overturning moments and lateral forces are generated on rolling stock. The pressure varies with the square of the wind velocity,and the inclination, area, and aerodynamics of the structure. The worst railway bridge disaster solely attributable to wind loading is well-documented: the collapse on 28 December 1879 at Dundee, Scotland, of the 13 high girders of first Tay Bridge into the Firth of Tay. They fell when the evening northbound mail train from Burntisland (Burntisland is more or less opposite Edinburgh on the north side of the Firth of Forth) presented a form of sail to the wind, enough to tip the final balance in a particular heavy gust. The inquiry found that the first Tay Bridge had been badly designed, with no allowance at all for lateral wind loading. It had also been badly built with gimcrack pier castings containing holes plugged with beeswax and painted over, and badly maintained, with loose bolts and broken lugs for which there were no proper inspection, reporting and repair procedures. Around 82 lives wore lost,the number arrived at by counting collected tickets, for nobody knew how many people were aboard. The newly knighted engineer Sir Thomas Bouch was held responsible, and the event was immortalised by a local poet named McGonigal in some of the worst verse ever written in English, and in one of the best novels, Hatters Castle, by A J Cronin. SURPRISING. in the eady days of the American railways there were bridge disasters aplenty several much worse than the Tay. But no large-scale American accident seems to be attributed to a bridge being blown down. This is surprising, given the largo number of high and exposed timber trestles structures over deep ravines. Indeed,one 1867 report of such a trestle on a transcontinent route said it swayed so much under the trains that it had to be steadied by lateral guy ropes. In the earlier days, however, the wind did cause one of America's more exotic bridge falls, at Spuyten Duyvil, New York, on 14 January 1856. At low tide, the wind drove broken ice across the frozen Hudson and piled it under the Spuyten Duyvil bridge. When the tide rose, the floating ice lifted one span off the bridgo and displaced it. The first train to appear was a double-headed mail (yes, double-heading was needed even on Sundays way back in 1856). The pilot engine got over safely, but the train engine and three cars wore dropped into cold shallow water,killing two enginemen. Ice drift is one form of wind hazard that need not concern Australian railways - but what if the ice had been a empty barge? Stability in high lateral winds was one of the reasons why military engineers recommended the 1676mm (5ft 6in) broad gauge for all strategic railways in India. But it seems that standard medium and even narrow-gaugo trains are really quite hard to blow off the rails. The main exceptions have involved narrow-gaugo equipment in very exposed positions. On the North Island of New Zealand, palisades were put up to protect trains after one was blown off the old Fell system incline up the Rimutaka Ranges, north-wast of windy Wellington. On the Midland Line from Christchurch to Greymouth on the South Island, steel palisades still protect the trains crossing the great high steel viaduct near Staircase (see picture on page 61 of April-May 1988 Network). On the 914mm gauge Burtonport Extension line of the Londonderry and Lough Swilly Railway in County Donegal, Eire, since closed, a train was blown off the rails on the relatively low but exposed Owencarrow viaduct on 30 January 1925. One car went over the edge, costing four lives. MAJOR HAZARD. But it is not necessary for the wind to drop a bridge or derail a vehicle to create a major railway hazard. Ill-secured parked wagons have been caught by gusts Of wind and blown into motion. And on lines without siding trap points or scotch blocks, this can create a full- scale main-line collision hazard. Lest anyone think that the trap points common in Australia would prevent these high jinks consider what happened on the sidings at Chelford, Cheshire, on the London and North Western Railway on 22 December 1894. The wind blew a high-sided wagon into motion, causing a violent collision with other wagons being shunted. This collision derailed them, and the pile-up fouled the loading gauge of the up main line just before the arrival of the double-headed 1215 express from Manchester to Crewe. The pile-up demolished the wagons, the express and the signalbox in front of the signalman, who, like a good railwayman, stuck to his post to morse out the "Obstruction Danger". Fourteen people ware killed. On 27 March 1884 a string of loaded boxcars parked at Akron, Colorado, ware blown into motion eastwards down the Buffington Company's main line. The telegraph wires hummed and the single track was cleared as the unmanned engineless train bounded along the main line, clattering over the points, through the stations, and passing trains, duly flagged through by an astonished railway staff. Finally, at Benkelman, Nebraska, an engine was found pointing the right way and despatched to chase the biowaways. Eight kilomatres further east,it coupled-up on the move and slowed the errant boxcars to a safe stop. The wind had blown them across the prairie for 154km or in Australian terms, half-way from Melbourne to Albury. This is somathing worth remembering, we suggest, when wondering whether to tighten-up that handbrake one more click of the ratchet. This Artical is Taken from "NETWORK Railways of Australia Vol26 No1" 1989 , Thanks.. Please notify me of any mistakes or Ommisions .