SOME CAUSES OF FAILURE IN WATER MAINS AND PIPES.

SOME CAUSES OF FAILURE IN WATER MAINS AND PIPES.

(Contludrd from last week.)

BREAKAGE of mains from water ram and hammer is much more frequent than is generally supposed. On pumping mains we expect an amount of concussion and variation; but, although we are well acquainted with the fact that the hydraulic ram will raise water, and, therefore, exact a pressure ten times that of its head, we rarely consider that on our service pipes we have apparatus that may put a strain far above what the head of water would lead us to suppose. The abolition of the old plug tap removed one source of concussion; but the ordinary self-closing “push” valves, so much in evidence at drinking fountains, will, especially at high pressure, put a very undesirable strain on pipes. A tap of this description with a three-sixteenth-inch waterway was attached to a twelve-inch main, on which there was a steady head of forty-six feet, a pressure gauge being fixed on a branch pipe too yards away. Pulsations were noticed on the gauge, at times the needle rising to ioo feet, and on rtioving the gauge to the service close to the tap 320 feet were shown, while a fire hydrant on the branch carelessly shut burst the pipe, the recorder showing at least 500 feet head. Hydraulic organ blowers working with a head not exceeding too feet, and probably much less, burst a connection which for days had stood the test pressure of 550 pounds to the square inch. Unbalanced ball and relief valves are also sinners in this respect—the “dancing” of the latter sometimes causing the burst it is fixed to prevent; and the sudden opening of the sluice-.valve admitting water to a district, especially where services are few, is sometimes the cause of a leak from the shifting of the lead in the joints; but cases of splitting of the main have also occurred.

Corrosion and incrustation are the next failures to be mentioned. The peculiar behavior of water on mains is often perplexing to the engineer, As a general rule*, soft surface waters, flowing from drainage areas, corrode and incrust the main by oxidation, diminishing both their strength and waterway, while hard waters derived from springs in a limestone district incrust by depositing crystals of lime; but we find soft waters which incrust, but do not oxidize, and hard waters which oxidize, but do not deposit lime crystals The same water will have no effect on one particular part of the main while it attacks another. On one pipe, coated with Dr. Angus Smith’s composition, the interior was found twentyfive years after as bright as the day on which it was dipped; while another, a few hundred yards away, in one-tenth of the time the water had dissolved the solution, and great wart-like carbuncles of iron oxide had formed on the interior. What is the cause of this anomaly ? Analysis of the rusty deposit, which, by the way, generally adheres to the upper part of the main, show s it to consist mainly of ferricoxide with some matter of vegetable origin, while calcareous deposits are chiefly calcium carbonate.

The ordinary rusting of iron, so familiar to all, depends on something more than exposure to an atmosphere containing oxygen— for specimens of iron broken to show the fracture keep their brightness in some situations for years. The presence of other substances, notably ammonia and carbonic acid gas, seems necessary for the oxidation of iron, and under certain circumstances this action is intensified. Iron placed in water in a warm position and carbonic acid gas introduced is speedily attacker!, the result being first an insoluble ferrous carbonate, which is rendered soluble by a further addition of carbonic acid gas, and the soluble salt acting on the water decomposes it, the iron taking up an additional quantity of oxygen which is sufficient to turn it into insoluble ferric oxide; the hydrogen and carbonic acid gas being given off and the acid again acting on the iron and again being decomposed by the water, it will be seen that a small quantity of carbonic acid gas will rust any amount of iron. Examinations of our mains will often disclose an identical process going on. The water leaves the service reservoir often charged with carbonic acid, and free oxygen travels swiftly under pressure towards the centre of distribution, but more slowly towards the end of the service mains, where pressures are variable at different times of the day. When the pressure is reduced, some of the dissolved gases are set free, and, rising to the top of the main, are strongly attracted by it; or it may be that the pipes are laid shallow-in summer the water is heated, and the gases being less soluble in hot than cold water,they are given off and similarly attracted, and the conversion of the iron of the main into ferrous carbonate is begun. Coating with Angus Smith’s solution does not always protect the main, this composition being soluble with some waters when they attain a temperature over fifly-five degrees Fahr., giving the water a strong pitchy taste. Mains rising to a dead end are more liable to this form of incrustation than those laid on a falling gradient, and the quantity passed through the pipe has also some influence on this form of incrustation, those through which little water passes being more liable to choke than those through which a large quantity is regularly delivered. Where the deposit is carbonate of lime it is mainly due to a similar cause, the dissociation of the carbonic acid turning the soluble calcium salt into an insoluble one. As a last cause of failure we may take the corrosion of cast iron mains from the outside—an experience happily of a somewhat local nature.

Occasionally this form of failure is met with in mains which have been down only a year or two, and a burst taking place on baring the main it is found that for some yards the metallic iron has disappeared, leaving sometimes only an oxide, but often a substance resembling graphite in its place,which when first uncovered cuts soft,like cheese,but hardens on exposure to the air. Sometimes on exposing a main of this description to the air the metal will get too hot to be held in the hand At another time the main rusts slightly on the exterior, and will attract a concrete like substance, which fastens like a rock round the pipe; and there are occasions when a three-inch pipe will measure ten inches across, the main being apparently little affected by this increase in bulk. Inquiry will generally show’ that in the case of speedy deterioration of mains where the iron is ” transmuted,” so to speak, this generally occurs in new districts where the streets have to be made up with refuse from chemical works containirg sulphurorchlorine,which develop acid properties, attacking the iron. The same effect may take place more gradually,but not th*e less surely in other places often away from towns and chemical works. Even in ordinary made ground in towns, especially where the sewerage system is, or has been defective, mains laid near sewers will suffer, the sulphureted hydrogen and chlorides in the soil being probably the cause.

As regards consumers’ lead and wrought-iron services, time will not allow more than a reference to some of the lead pipes at the present day, which are hard, and break with a crystalline fracture on any attempt being made to bend them. Only an approach to the ideal water main can yet be made. A cast iron pipe made of a white, hard pig in which the crystals are small seems to lie the strongest and at the same time the material which resists corrosion best, although more liable to crack if carelessly thrown down. These pipes should be laid in warm weather if possible, so that no allowance need be made for expansion, but if laid in frosty weather a lead ring or a papier mache wad should be inserted in the socket and the spigot brought up against it. Only enough spun yarn should be inserted to prevent the lead getting inside the pipe, and forced well back with the yarning iron, the rest of the pouch being filled with lead at a temperature sufficient to char the yarn, enough head being kept on the lead to keep the upper part of the joint full and cause any scum to rise to the top of the runner, the lead being then well set back nearly flush with the socket, and except in the smallest pipes every joint should be run in situ. According to the size of main and nature of the roadway, the trench should be deep enough to leave from two feet six inches to three feet six inches of cover on the top of the main; small pipes requiring the maximum. If unaccompanied by snow, frost will go down under a pavement to a depth of three feet six inches. On a macadamized road it does not descend so far; on graveled footpaths it descends still less; and on ploughed land it descends a very few inches. Heat in the summer months will descend to a far greater depth than cold in winter months, fend the temperature of the water at consumers’ tap is more affected thereby. Service mains should be well provided with stop valves.and all dead ends avoided as far as possible,so that, when the pipes are under repair, a supply can be given on either side to all except that particular section under repair. Larger mains are required on works to day than were required to supply similar populations twenty years ago, the legitimate consumption per head being much more in these days of advanced sanitary attainments—and more particularly in a corporate supply no street pipe should be laid less than three inches in diameter. It would, of course, be foolish to connect a three inch extension to a two inch or one and one half inch supply; but apart from the requirements for fire service,which are generally ignored by companies, the economy of small services is doubtful; incrustation is more fatal to a small, than a large main, the effective waterways being more proportionately decreased, and as below three-inch pipes are cast in sixfoot instead of nine-foot lengths the number of joints is in creased one third; the cost of extra labor going somewhat to pay for extra material. Unbalanced ball-valves and other ap. paratus causing concussion when closing should be interdicted, and all sluice-valves and fire hydrants closed gently. Frequent scouring at the ends is the best preventive of incrustatation; pipes with a good (low through them incrusting less than where the water is nearly stagnant. Where the distance between the service reservoir and the town exceeds two or three miles, a duplicate main, capable of conveying sixty per cent, of the maximum supply should be laid, no connection being fixed on this until it reaches the centre of distribution. This is somewhat costly in many cases; but one main, twelve or fourteen miles in length, the breakage of which would interrupt the supply of a large community, is not enough to depend on by itself. Further, the abolition of storage cisterns and intermittent supply has tried the capacity of existing works—in some cases half the daily supp’y is required between the hours of 7:30 and 11:30 a. m.; b it few mains are equal to deliver this quantity, and the higher localities suffer from a want of pressure. While waste from any source should be firmly put down, every facility for the legitimate use of water should be given, and under no circumstances should a necessity of life be measured out with a stinted, niggardly hand.

SOME CAUSES OF FAILURE IN WATER MAINS AND PIPES.

0

SOME CAUSES OF FAILURE IN WATER MAINS AND PIPES.

ALTHOUGH examples of lead service mains which delivered water to some districts in Rome and Pompeii 2,000 years ago may be seen in the Gregorian Museum at Rome, the Royal Museum of Naples, and other places,and in this country one occasionally comes across a lead pipe which has acted as a carrier of water for 200 or 300 years, mains for carrying water under pressure may be taken as a result of the age of iron, and do not date back farther than the century. The wooden predecessors of our cast iron mains are now and then unearthed, generally in such a state as to cause the most phlegmatic resident engineer to be thankful that he lives in the nineteenth, and not in the eighteenth century. What with “spigot and socket joints,” “ casting vertically in dry sand,” testing by hydraulic pressure, and coating with Dr. Angus Smith’s solution, the question of deterioration of service mains may, as a rule, be a very secondary consideration to the designer of water supplies.

While probably most of us, as our area of distribution has increased, have been obliged to enlarge, or, better still, duplicate a line of service main to a given centre, few have the misfortune to be called upon to lay a new line of cast iron pipes in place of an old one, owing to the lattter being worn out and not worth repair. This experience unfortunately does not extend to consumers’ services,and we can all recall dozens of instances where, on the faulty pipe being handled, it has snapped easier than the proverbial carrot, or has been corroded through; a further examination condemns the whole length, which is replaced perhaps by a similar pipe, on which the action begins de novo. Failures in service pipes are the bane of the resident engineer’s existence. In average works probably nine-tenths of the total waste may be debited to their account; and,as such waste often finds its way into a sewer.it is difficult to detect and may exist in works which have a thorough system of inspection—for the leakage may be on a disused service pipe or too small to inconvenience any of the consumers, and the inspector trusting to his eyes alone the defect may remain unnoticed.

The scientific design of a service main docs not seem to enter into the consideration of the engineer to day so much as it did fifty years ago. Although the discharge with a given head is well understood, the pipes are rarely enlarged on a flat and diminished on a steep gradient, and more rarely still increased or diminished in thickness as the pressure to which they are exposed is greater or less. The reason of this is not far to seek. With many foundries cast iron pipes have become specialties, standard sizes, lengths, and weights being cast suitable for any ordinary pressure in water works practice. Engineers who have taken the trouble to calculate the thickness of cast iron required to resist the greatest head on their district have found, especially in small pipes, that even with a margin of safety of four and a low tensile strength, for low pressures it worked out to a thickness so small as to be impracticable; and pipes strong enough to resist bursting pressures are generally doubly so as to resist any end strain which would tear them asunder. The requirements of enough metal on the pipes to take a thread for the insertion of ferrules and to resist the transverse strains of earth movements and heavy traffic probably had more to do with the standard thickness of cast iron mains than any fear of their inability to withstand the pressures; ami it may,therefore,be taken as an axiom that no standard iron main well cast, of good material, and free from flaws, has given way from weakness in design; and yet every year cases come under our notice of failures of pipes— some apparently sound, others with flaws more or less easily traced. Even without the strict inspection which water undertakings provide at the pipe foundries when large quantities are required the contractors themselves see that the pipes that leave their works are cast uniformly and stand the hydraulic test imposed, not only to keep up their good name; but,knowing that any discovered faulty pipes will be returned, carriage forward, probably accompanied by a tersely sarcastic notification of its quality, it rarely happens that a faulty length, especially of the larger sizes,ever leaves the foundry, and the defects on the smaller pipes are mainly “cold shuts,” resembling a bad weld caused by insufficiently fluid metal and spongy sockets, both of which are somewhat difficult to detect when they show to the interior.

“Specials” from local foundries are often indifferent— “ double collars,” curiously enough, in the author’s experience, providing the greatest number of failures—and it is peculiarly aggravating when,having replaced a defective main, depriving a district of water for an hour or two to find the metal is porous, and a slight “ weep” necessitates the bursting of the collar and the undoing of the whole work. Bends (quarter) have sometimes cavities running diagonally across their centres which would remain unnoticed until a more obtuse angle than ninety degrees is required, the defect being shown on cutting the pipe. The branch-pipe, which is a combination of an angle, tee and bend, is occasionally a trouble from moving of the core. Violent shunting and careless unloading, especially in frosty weather, will also cause cracks in mains, which may remain unnoticed until the water is turned on.

Coming to joints: The spigot and socket joint,in which a strand of rope yam is inserted and the remaining space filled with molten lead, which on cooling is “ set back ” with a caulking tool, is the favorite, and, fairly made, this joint will stand any pressure likely to be found on water undertakings; indeed,they have been used on hydraulic mains; but above 500 pounds the lead has a tendency to work out. In mains laid by contract, and insufficiently inspected, it is sometimes found that the bulks of lead and yarn have changed places, the pouch being mainly filled with yarn, which forms the nucleus of future troubles. The yarn rots and finds its way into the mains, to the great deterioration of the water supplied, and on mains which have been carelessly laid one may occasionally see yards and yards of spun yarn pass out of the hydrants during scouring, while the space in the pouch forms a starting point for shifting lead. To prevent this, some engineers use a ring of lead, which is first caulked back into the socket pouch and the ordinary lead joint run and set back afterwards. Turned and bored joints made with tallow, red lead, or Portland cement are favorites with some engineers, and have the merit of being quickly and easily made but with the disadvantage of being flex ible, and, therefore, unsuitable for slight changes in level and direction as well as expansion and contraction where there are extreme variations of temperature. On the Boston works the service mains from the reservoir to the town, twelve inches in diameter are jointed for twelve miles in length with wood wedges driven into the socket in place of lead, and although they have been in use nearly fifty years have not cost a sovereign in repairs, and to all appearance will stand another fifty years.

Coming to damage caused by violence, there are many local circumstances which vary the liability to damage from this source. In localities where mining operations are carried on, the drawing of joints and breakage of mains from the irregular subsidence of the ground is the cause of continual worry and expense to water undertakings quite unknown to other districts, whose only approach to similar delights is the occasional unskilful execution of sewerage works in their district causing the earth to move. Then extraordinary traffic, such as traction engines using the roads, or, worse still, steam-rollers, makes them play havoc with our smaller service mains, especially in new localities where the subsoil has not got consolidated. Again, laborers taking out a trench for gas or sewers have a wonderful faculty of driving their pick through a water pipe, even when its exact position has been pointed out to them a few minutes before; but the loss occasioned by all these causes put together is insignificant when compared with that caused by the almost annual advent of frosty weather. The damage caused by water actually freezing, in and bursting the pipes is, as a rule, confined to consumers’ services. But some of us have another kind of failure indirectly caused by frost, which in some towns causes a great amount of damage, while in others they have rarely a case. Where water has frozen and expanded enough to burst the pipe, the failure is invariably of that longitudinal form term a “split.” which shows directly the thaw comes; but in the cases under notice the pipe breaks across, generally much more regularly than it could be cut by a chisel. The time of breakage varies, in some cases being confined to the time of severe frost; in others it occurs sometimes at the time of the frost, but generally fifteen to twentyone days after the thaw. The smaller-sized mains are the ones most affected,and the composition of the subsoil seems to be the principal factor, the temperature of the atmosphere and water in the main being also considerable factors in this matter. In ltoslon up to January. 1891, these breakages invariably occuired about a fortnight after the thaw: but in that year the reservoir, being unusually low, the breakages occurred during the pre valance of the frost,and reached the maximum of any year recorded—while in January, 1895,with all the terrib’e weather experienced, the breakages from this cause were far below the average. Observations have shown that, with snow on the ground and a fairly full storage reservoir, the temperature of the water in the service tank, even when it and the filters were covered with ice, did not fall below thirty-six degrees Fahr., which increased three degrees and four degrees on its way to the town. After the thaw, the melted snow brought the water to nearly freezing point, the temperature in the town mains falling to thirty-three degrees Fahr., at which point the subsoil was probably affected, and, moving, caused the fracture. In January, 1891, a long drought and frost had practically emptied the reservoir, the water left fell rapidly in temperature, and after a month’s frost the low temperature was noted in the town; but in 1895, the storage reset voir was full, and, with an absence of snow, the temperature in the main did not fall below thirty-six degrees Fahr.

(To be continued).