THE COLLIERY GUARDIAN AND JOURNAL OF THE COAL AND IRON TRADES. Vol. CXV. FRIDAY, JUNE 28, 1918. No. 3000. STEAM PIPE EXPLOSIONS. By EDWARD INGHAM, A.M.I.Mech.E. In looking through some of the reports on boiler explosions issued by the Marine Department of the Board of Trade, the writer has been struck with the comparatively large number of explosions which have taken place at collieries and iron works. Some of these have occurred in connection with winding engines, others with steam boilers and economisers, but the great majority appear to have been steam pipe explosions. At most collieries and iron works large ranges of steam pipes are to be found, and it is a fact that in many cases the pipes are not designed to the best advantage, the result being that a certain amount of danger in working is incurred. A brief consideration of steam pipe design and arrangement will therefore no doubt be instructive, and may help to reduce the number of explosions which occur. The principal factors responsible for steam pipe explosions are three in number, viz., water hammer, expansive movements and vibration, of which the first- named is by far the most important. Before proceeding to discuss these, it should be explained that steam pipes are generally made to with- stand a pressure many times as great as the normal steam pressure, so that inherent weakness is, as a rule, out of the question. It is thus quite evident that, since explosions occur so frequently, forces of excessive magnitude must be brought into play under working conditions. Water Hammer. Water hammer is the action set up when a body of water which has been allowed to collect in the pipes is propelled violently forward by the steam, so that it strikes the metal walls of the pipes or the valves with great force. When the flow of steam is shut off by closing the valve, the steam remaining in the pipes condenses, and a quantity of water collects. If now the valve be opened again, the entering steam, on coming in contact with the accumulated water, rapidly condenses, and a partial vacuum is formed. The steam behind then rushes Qnwards at high velocity and propels the water forward, until the latter strikes an opposing metal wall, involving risk of fracture and explosion. The obvious remedy for water hammer is to avoid accumulations of water in the pipes, and it is there- fore of prime importance that in designing a steam pipe range one should endeavour to arrange the pipes in such a way that water cannot collect at any part. The ideal arrangement is that in which the pipes have a gradual and continuous fall from boilers to engines, but unfortunately such an arrangement is often im- possible because, owing to changes of level, bends and dips in the pipes are necessitated, with the result that “ collecting lengths” are formed. It is, however, a fact that in .many steam pipe ranges the number of these collecting lengths might be reduced by careful design. There is, for example, much room for improvement in the design of the pipes connecting the boilers with the main range. It is a common practice to bolt the stop valve directly on to the mounting block of the boiler, and to connect the valve with the main range by means of a bend pipe (see fig. 1). Clearly, with this arrangement, water of condensation can, collect above the valve when the boiler is shut off by closing the valve. The diffi- culty may be easily overcome by bolting a stand pipe on to the mounting block, so that the valve is lifted up to the levels of the main range (see fig. 2). Other similar examples might be given, but this will suffice. Where it is impossible to avoid water pockets, careful attention must be given to the question of draining. Efficient drains should be fitted at all parts where water can collect, and these should preferably be connected with automatic steam traps, so that the water of condensation will be drained off as fast as formed. There is always the possibility that an attendant may forget to open the drains before turning steam into the pipes, and many cases of water hammer have been the result of such forgetfulness. It is important that the drains should be of ample size and be cleaned out periodically, since they are liable to become partially choked after the time.' A recent explosion at a colliery near Nottingham was the result of a partially choked drain. The practice of opening drains whilst there is pres- sure in the pipes is one which should be avoided. The belief that no danger is incurred in doing this is erroneous. There may be no risk in draining the water from a vertical pipe under pressure, provided the pipe is not connected with a long horizontal length below, but otherwise considerable risk is incurred, and this is evidenced by the fact that in several instances attendants have been killed or badly scalded whilst in the act of draining a pipe under pressure. In the case of an explosion which occurred at a colliery near Derby, the attendant opened a drain cock to run off the water from a vertical length prior to opening the junction valve. The drain, however, was partially choked, so that the water was not got rid of before steam was turned into the pipes. Drain- ing continued very slowly under pressure until the level of the water reached a horizontal length of piping. A large surface of comparatively cool water was then exposed to the steam, water hammer being the result. To minimise the risk of water hammer, it is always advisable to open the junction valve very gradually. It is not difficult to realise that the sudden admis- Fig. 1 Fig. 2. sion of steam into the pipes is far more liable to set the accumulated water into violent motion than a very gradual admission. Expansive Movements. The amount of expansion which takes place when steam is turned into cold pipes depends upon a number of factors, but for saturated steam of ordinary pres- sures, say 100 to 200 lb. per sq. in., it is between 2J and 3 in. per 100 ft. length of iron or steel piping. With superheated steam the expansion is much greater than this, and may, indeed, be as much as 5 in. when there is a high degree of superheat. Since the atomic forces of expansion are practically irresistible, it is evident that if safety in working is to be ensured, suitable provision must be made for accommodating the expansive movements. From a theoretical point of view, the best arrange- ment for this purpose is a gland expansion joint, which consists of a sleeve sliding through a stuffing box and gland, but in practice these joints frequently prove to be unreliable. To ensure reliability two important conditions must be fulfilled. The first is that the pipes should be erected in true alignment, and the second, that the gland packing should be frequently renewed. When the former condition is not observed, the sleeve of the joint cannot slide freely in the gland, and sticking ensues. Much the same thing occurs when the packing is not frequently renewed. After a time it becomes hard, and to pre- vent leakage, the gland must then be screwed up tightly. This is liable to cause binding, in which case the expansive movements cannot be taken up by the joint, but must be accommodated in some other way, probably by bending and consequent straining of the pipe line. It is important that all gland expansion joints should be provided with guard bolts, to obviate the possibility of the ends of the pipes being drawn out of the joint. Numerous scalding accidents have been caused in this way. Instead of expansion joints, spring bends of large radius are commonly used for taking up expansive movements in steam pipes. Two common forms are illustrated in figs 3 and 4. The bends are placed at intervals in the pipe line, and should be arranged so that water of condensation cannot collect in them. In designing a p^pe range it is often possible to provide against expansive movements by a suitable arrangement of the pipes. Thus, by inserting spring lengths at various parts, the pipe range, if not of great length, may be rendered sufficiently elastic to take up the expansion without the assistance of gland expansion joints and spring bends. Vibration. This is a common trouble with steam pipes, and is generally the result of pulsations of the steam in motion. In some cases an improperly balanced engine is the cause, and should the pulsations of the engine happen to synchronise with the natural period of vibration of the pipes, the latter may vibrate to a serious or even a dangerous extent. The general effect of vibration is to cause fatigue, or gradual weakening of the pipe material, and in serious cases the pipes eventually become too weak to withstand the pressure, and an explosion is the result. To prevent vibration is is necessary to fix the pipes at certain points, so that they cannot move trans- versely, but care must be taken that there is no inter- ference with the freedom of longitudinal movement necessary for expansion and contraction. Probably the best arrangement is to anchor the pipes securely to a wall at intervals, and support them on adjustable rollers fixed in cast iron brackets. With this arrangement the pipes are prevented from moving transversely, but are free to move longitudinally. Material for Steam Pipes. The question of which is the best material for steam pipes is a most important consideration. Formerly cast iron was generally used, but experience has shown that although this material is more or less satisfactory for low steam pressures, it is quite un- suitable for high pressures and for superheated steam. Not only is it brittle, but after continued exposure to the action of high temperature steam, it becomes weakened to a serious extent. Fig. 3. Fig. 4. Hence cast iron has of recent years been superseded by wrought steel, which is more elastic, stronger and lighter. Cast iron has the advantage of cheapness, especially when there is a duplication of pipes, and for this reason will no doubt continue to be largely used for low pressures. Copper is an excellent material so far as flexibility is concerned, but is not recommended, being liable to deteriorate under the action of high temperature steam, and to become brittle if subjected to continuous vibration. Expansion bends and cushions are often made of copper, but it is important that these should be annealed every two or three years, so that in case the material has suffered deterioration or become brittle, it may be restored to its original condition.