TANK CAR SAFETY

[Music] [Music] Millions of miles of fast, efficient, and safe transportation are logged by the nation's freight trains Each year, Americans are so used to the trains rolling that people don't even seem to notice them. What the public does notice is the sudden news break, a train wreck, derailment, punctured tank cars carrying hazardous material, then an explosion and tragedy. After each headline grabbing incident, the question is asked, why isn't something being done about it? The answer is quite simple. There is something being done about it. to the problem was conducted from one end of the nation to the other utilizing both government and industry facilities. Both groups knew that the first step was to understand the anatomy of the accident to discover just what led to the chain reaction explosions that uniquely accompany tank car derailments. These researchers knew that America had been moving hazardous material over long distances by one of the most efficient and safest ways known in railroad tank cars. The transporting of flammable liquids such as propane is safe and uneventful under normal circumstances. However, the terrible results of the abnormal caused experts from both the public and private sectors to investigate just how these disasters came about. Investigators were on the scene of every accident, inspecting the wrecks, interviewing eyewitnesses, rebuilding the sequence of events. For the accidents themselves were not always major. It was when chain reaction explosions followed that there were disasters. Researchers in government and in private industry traced every clue in order to explain the circumstances leading up to the disasters. Finally, the investigators were able to determine the cause of the chain reaction explosions that follow tank car accidents. They found that if a pressurized tank carrying flammable compressed material is cracked, torn or punctured, its contents can be released in two characteristic modes depending on the temperature, pressure, size, and location of the hole. In the first mode, the liquid content escapes in a stream and sinks to the ground, forming a pool. Since it is frequently heavier than air, any sort of spark can ignite a fire. Often neighboring tank cars are subjected to intense heat. Thus, the chain reaction sequence is established. The violent rupture of a tank causes large sections of the car to rocket over 1/2 mile, and they often trail ignited liquid. The ground flash covers an area hundreds of feet in diameter. The blast wave has tremendous force, enough to break windows several miles away. The fireball rises on a thermal column, radiating heat severe enough to cause burns for distances up to 1,000 ft. Pieces of the exploding car can puncture other tank cars, spreading the chain reaction to cars far away from the original rupture tank car. Only the contents of one car need be spilled to begin a chain reaction, with each ensuing explosion capable of touching off more and more cars. This all happens swiftly without warning. It takes anywhere from 15 to 30 minutes to get equipment to the scene of the accident. By this time, the fires are so intense, little can be done. Most times, the explosions seem insidious, waiting until firefighters are on the scene to cost the most in lives and damage. The researchers knew that to prevent the thermally induced chain reaction explosions which sometimes accompany tank car accidents, a method must be found to protect even those tank cars undamaged in the accident from the devastating effects of both pool and torch fires. Research and testing was underway at White Sands, New Mexico, where a BRL team was setting up pool fire experiments at the Transportation Systems Center in Cambridge, Massachusetts. They were building mathematical models of tank car collisions that cause puncture. In Pueblo, Colorado, the transportation test center was preparing to test those theories. The special RPI AAR railroad tank car research and test project at the AAR research and test center in Chicago were testing flame retardant systems. The work at the White Sands Missile Range was undertaken to discover more about the effects of pool fires under undamaged tank cars that are part of the pattern of the deadly chain reaction. At first, several different types of tank car insulations were tested on 1/5th scale models. Instrumentation was connected at each end of the tank to measure temperatures and pressures inside. After all the testing on the 1/5th scale tank was completed, tests on actual tank cars were started to determine the effects of a pool fire on full scale cars. Working through the research arm of the Federal Railroad Administration, the BRL team began the testing on actual tank cars at Whit Sands. The tank cars were donated for this purpose by the Association of American Railroads and the Railway Progress Institute. A normal uninsulated test car was put on rails in a pit. It was just like the 22,000 tank cars that were rolling on America's rails. Fuel was fed through an underground pipe to form a fire pool under the car. 1 minute 40 seconds after the start of the fire, the pressure inside the tank car had reached the safety valve and relief pressure levels of 280 lb per square in and the first gas venting occurred. A second and third venting occurred soon after and from then on the internal pressure remained above the 280 lb per square in level and there was continuous venting. Observers estimated that the flame at times reached more than 250 ft in height and averaged about 100 to 150 ft. The internal temperature reached 1,72° F. The pressure peaked at 357 lb per square in which was 77 PSI over the relief valve set pressure. Taking a look at the conditions inside the tank car, we find where the flames contact the tank below the liquid level, that liquid acts to absorb heat and the tank metal below this point often remains at a safe temperature for a long period of time. However, as the liquid is heated and vaporizes, the liquid level within the tank drops, exposing greater areas of the tank metal to the effects of the flames. Flames contacting the tank above the liquid level, which is rapidly dropping, create temperatures within the metal high enough to weaken it. As the temperatures without cause the metal shell to weaken, the pressures from within the tank car also cause it to thin and tear. Although shown in slow motion, the pressure drops in an instant. Large quantities of the boiling liquid expand, vaporize, explode. After 24 and 1/2 minutes of exposure, the tank exploded violently into 128 identifiable pieces. Even the rail trucks were completely destroyed. That 24 and 1/2 minutes would normally have been just time enough for firefighters and inspectors to have gathered around the tank car. It would have been time to do the most damage in terms of human life and property. The safety pit was 100 ft wide, 150 ft long, with walls 26 ft high. Practically all of the tank was blown out of the pit. The furthest piece was located some 1300 ft away. The black coloring along the debris path indicated there was some directional propelling of parts and burning fuel involved in the explosion. Approximately 15,000 gallons of propane remained in the tank at the time of the explosion. Now we knew what actually went on in an uninsulated tank car in a pool fire environment. Although appearing identical to the first car, the second test car was sandblasted down to metal, primed, sprayed with 1/8 in thermal shield material, and finished with a weather resistant top coating. The insulating jacket outside the tank shell delays the heat induced weakening of the tank metal until the boiling liquid contents are fully vented. The first difference noted was that it took 15 and 12 minutes before venting occurred in the second car with thermal insulation. The uninsulated car exploded in 24 and 1/2 minutes. In the insulated car, the fire burned for 94 1/2 minutes before the rupture occurred. In the uninsulated car, approximately half of the 30,000 gallons of propane gas remained when it exploded, spewing that amount of ignited fluid along its destruction path. In the insulated test, 29,000 gallons burned off before the explosion occurred. Researchers estimated that if the car had withstood the pressure another 10 minutes, all the liquid fuel would have burned off. Without liquid propane in the tank, the chance of damage and personal injury from rocketing car parts is greatly reduced. In the uninsulated car, pressure went to 357 lb per square in. In the insulated car, pressure never exceeded 320 lb per square in. The insulated car ruptured into four identifiable pieces. Two small pieces were thrown approximately 500 ft out of the pit, but two major fragments remained in the pit together with the two rail trucks which were completely intact and in place on the rails. The results of these tests seem simple and clear. A tank car protected with thermal material could extend reaction time and prevent dreaded chain reaction. Because the extended venting time would permit all or most of the boiling liquid to escape, it would reduce the severity of the rupture. Through the research effort, a second mechanism of fire exposure was identified. In the second mode, the contents of the damaged car is projected in a high velocity stream on a neighboring car. This venting through small punctures or the relief valve of a derailed tank car sometimes has the effect of a blowtorrch on the car next to it. To fully investigate the effect of a concentrated torch fire on the tank car metal, the BRL researchers designed and constructed a torch simulator at the transportation test center in Pueblo. An uninsulated car that is subjected to a highly concentrated torch of intense heat can explode in less than 5 minutes, especially if the affected area is above the liquid line. This very drastic reduction in time is due to the concentrated high temperature and velocity of the torch on a relatively small area of the car where the liquid is not capable of absorbing the destructive effects of the heat on the shell. First, the blowtorrch effect was concentrated on material plates. Unccoated plates were compared with those coated with thermal material and those additionally covered by a steel outer jacket. The insulation protects the car metal from the more intense torch which is fueled from the punctured car. The fuel level in the punctured car lowers as the torch burns until it is unable to sustain the torching effect to ensure nothing was overlooked. Tests were performed with a fire hose concentrated on the torch plate. This simulated conditions when firefighters were present. The researchers concluded that the available fuel could sustain the destructive torch for up to 30 minutes. The challenge, protect the car metal for more than 30 minutes. The challenge was met and thermal protection systems were found that protect the car metal for more than 30 minutes. The tests were then transferred to real tank cars. In these torch tests, as in the pool fire tests, whether on scale models or actual tank cars, the results were always the same. Thermal protection, whether by coating or insulated jacket, reduce car temperature, which could extend reaction time and prevent chain reaction explosions with their potential for severe damage and personal injury. Tank car fires and resultant explosions are initiated after the cars have been punctured or cracked in accidents. More research and testing had been taking place to find ways to avoid the initial spilling of fuel. From the mathematical model of an accident that they had built, researchers in Cambridge theorized as to what caused some of these puncture type accidents. At the transportation test center, these theories were tested on actual rolling stock. Test configurations were designed to subject tank cars to situations causing impact at the ends or heads of cars. Although what you will be seeing will be slow motion film, the actual speeds were not high. They were between 6 and 19 mph. An empty hopper car was placed near a row of tank cars loaded with water. Another water-loaded tank car was sent down the track at varying speeds. The space between the hopper car and the stationary tank car is 3 1/2 ft. for this test. Spacing like this is possible when cars are humped or switched. When cars are not coupled, it sets up some of the most severe conditions upon collision. The center of gravity is indicated on the hopper car. On impact, the force of the striking tank car is below the center of gravity of the stationary hopper car. This horizontal force causes the impacted hopper car to move forward and at the same time to rotate, causing the far end of the car to pitch up. [Music] The coupler of the impacted car rises above the coupler of the tank car next to it and punctures the head. A series of extensive tests were conducted over a period of months. These tests demonstrated the various mechanisms causing override. These mechanisms working singly or in combination could lead to leakage of hazardous material under actual conditions. This leakage is a common source of fuel for some of the fires and chain reaction explosions that follow tank car accidents. Once the anatomy of the accident was understood, tests were conducted to evaluate the effectiveness of a countermeasure system consisting of two components to prevent tank car puncher. These two components were headshields and shelf couplers. These components were tested singly and in concert over a series of tests. Head shields can be trapezoidal shaped steel of 1/2- in thickness that are placed at each end of the tank car. These shields were designed to absorb impact energy and disperse the force of the thrusting coupler. The weight of the cars is the same as the earlier tests and the separation distance is 3 1/2 ft, but the speed is even higher. This created an even more severe testing situation than those that caused puncture. The mechanisms work in the same fashion. The car rises at the far end and thrusts into the adjoining tank car, but this time a part of the energy of the hopper car is spent crashing against the steel head shield, which spreads the load as it deforms against the tank car head, preventing puncture. Under certain combinations of circumstances, knifing couplers rode above the head shields to puncture the tank. This would not have happened if the couplers had not separated. This emphasized the need of the other component of the countermeasure system, double shelf couplers. Heavy steel shelves were added to the standard eoupler. It was anticipated that these shelves would prevent the separation of the couplers when one was being forced upward. The effectiveness of the double shelf couplers were put to the test at Pueblo. The design of the double shelf coupler was conceived to prevent the separation of couplers caused by the upward motion. As the colliding tank car crashes into the hopper car, we'll concentrate on the coupling of the hopper car and the stationary tank car. The same upward motion occurs under the similar conditions as in the earlier tests, but the added top shelf prevents the couplers from separating. Another advantage in the use of the double shelf coupler was demonstrated during the testing. Vertical forces acting on the flange of the double shelf coupler cause a large bending stress on the shank. This can result in a shearing off of the shank. Now the same force of impact that would have caused puncture is distributed over a larger surface on the tank car head rather than being concentrated in a small area. The double shelf couplers and 1/2 in head shields were each a component of a countermeasure system to reduce accidents. Shelf couplers helping in the prevention of coupler separation and head shields adding protection when couplers do not engage. Together they were effective in the testing at Pueblo. The work going on across the nation with the full cooperation between governmental agencies and private railroads and shippers has produced solid results. The use of thermal insulation was proven to be the key factor in retarding explosion time and minimizing damage. Add to this the puncture countermeasure system with its component parts, head shields, and double shelf couplers that prevent the fuel release needed to feed a pool or torch fire. The testing has shown there is an excellent chance of averting disaster. The need for such a combination of deterrence was so apparent to both the public and private sectors of railroading that changes were started almost immediately after the test results were known. As a result of the coordinated research accomplishments, a petition was voluntarily submitted by industry and the federal regulation was quickly promulgated by the Department of Transportation's Materials Transportation Bureau requiring the outfitting of thermal shields, head shields, and double shelf couplers on the existing fleet of tank cars. We have seen how research can be truly responsive to safety needs and achieve practical results. Cooperation and mutual understanding between private industry and government agencies can bring about giant steps forward in safety for the men who operate this vital link in our nation's transportation system and the communities it serves.