LEAKAGE FROM LEAD JOINTS IN CAST IRON WATER PIPES*
This subject deals with one phase of a “conservation” movement which has for years received careful consideration by many officials and engineers who are responsible for the security and distribution of our public water supplies. During the last few exceedingly dry years prior to the present season, when the margin between available supply and consumption has been rapidly narrowing, special attention has been given to the elimination of waste as a means of increasing this margin. Many towns and cities have found that by systematically reducing the amount of water running to waste, their present supplies could be made ample for their needs. By various systems of metering mains large leaks have been located and repaired, the installation of meters on house connections has largely stopped unnecessary losses taking place at plumbing fixtures and appreciable savings have been made. With the development of the pitometer into practicable form, additional impetus has been given to the locating of leaks in mains and extensive surveys have been made in New York, Philadelphia, Washington, Chicago, Milwaukee and other cities, the results of which are matters of record. The studies made have shown very large losses from three principal causes: (a) Leaking of underground mains: (b) Broken or abandoned service pipes; (c) Defective plumbing fixtures. Of these causes, the third is being corrected quite satisfactorily by metering. The first and second are more difficult to reach. In old systems it frequently happens that large leaks occurring in porous soil, under streams or near sewers may continue for years without giving any surface indications and thus escape the attention of the unthinking or careless official. Usually it is feasible at some labor and expense to locate and stop these larger wastes, but the smaller leaks occurring at imperfectly made lead joints or at minute cracks in the pipe represent a class of losses generally felt to be incurable. It is more especially with these types of leakage and some observations regarding the amount lost in this way that this paper deals. They rarely cause much concern to the water works man, although the total volume lost in a year is often very large. In an effort to determine the volume of leakage from underground mains, some officials have simply subtracted the quantities taken out from those fed into a system, but this has generally not been very successful because of the large number of assumptions or guesses which have to be made. Of course, no conclusions, even approximately correct, are oossible except where practically 100 per cent. of the service supplied is metered. The best performance along this line which has come to the writer s attention is that in the city of Cleveland, Ohio, where, under the administration of Dr. E. W. Bemis in 1909. all but 7 per cent. of the supply delivered into the mains was accounted for, a loss equal to only 6.3 gallons per capita per day. The same year the Water Registrar of Worcester, Mass., reported that all but 19 per cent., or 11.5 gallons per capita, of the supplies metered into the mains of that city was accounted for. In the several suburban communities of the Metropolitan District surrounding the city of Boston which are 100 per cent, metered, about 65 per cent, of the supplies entering the systems is metered out or otherwise accounted for. In an attempt to determine with some accuracy what is a proper allowance for this incurable waste, Mr. John R. Freeman secured data from several New England cities for his report on the water supply for the city of New York, dated March, 1900, to show the consumption during early morning hours which was assumed to be made up largely of leakage from mains. On a gallons per capita per day basis, these figures were: Fall River, 10; Woonsocket, 8; Boston, 10; Providence, 20; Milton, Mass., 3. It is not correct to assume that all of the leakage in the instances cited takes place at lead joints and cracked pipe, for there is an appreciable contribution from defective service pipes, hydrant valves, blow-off gates, leaky packing at gate valve stems, etc. While, therefore, the “gallons per capita per day” basis is well adapted to represent the total underground leakage, a more appropriate unit in which to express the leakage at joints seems to be “gallons per day per lineal foot of lead joint,” and in the cases quoted below, this standard has been used. In the writer’s experience with the Factory Mutual Fire Insurance Cos., questions have frequently arisen between the managers of corporations insured by them and public water works officials as to what becomes of the water passing from public mains into private fire protection yard systems. In such cases the water works official is usually well satisfied to install a meter on the main feeder and charge the consumer with the registered quantity of water, allowing him to use it for whatever purpose he desires, while the attitude of the insurance companies has been that if all domestic and manufacturing water is used from a separate metered line, the interests of the community are not jeopardized by omitting meters from the fire service lines, provided, of course, that these are used solely for this purpose. In other words, the water for extinguishing fire in protected plants should be as freely supplied and with as little obstruction to flow as it would be if taken from street fire hydrants. The arguments on both sides of this proposition are so well known that they will not be repeated. A table shows what general experience confirms that bottle tightness in lead joints of cast-iron water mains is an impossibility. So long as pipe lines are built in this way and from our present knowledge there seems no better method for general practice, leakage must be reckoned with. The oroblem then is to reduce it to a minimum. In this endeavor two general lines of action are possible—first, the substitution of some more satisfactory material for lead; second, an improvement in the design and construction of the lead joint. Under the first heading, there has already been produced a substance called “Leadite,” which has been quite extensively used. Whether this will solve the leakage problem and become the universally recommended material for cast-iron pipe joints is still an open question. Undoubtedly the papers to be presented at this convention will shed much light on the subject. Under the second heading, a general stiffening of specifications under which pipe is laid, with more careful supervision of joint-making, laying in deeper trenches to reduce expansion and contraction, mechanical injury, etc., and rigid tests for leakage after completion would have a tendency to produce the tighter work. The standard of tightness which is to be regarded as acceptable will be subject to much difference of opinion, depending upon the point of view. If leakage on new work can be limited to one gallon per day per foot of joint under 100 pounds pressure, the performance can be called good and acceptable. It would seem that below this limit, thought and energv might better be expended in tracing other forms of waste than that at lead joints.
*Abstract of paper read at New England Water Works Association Convention, New York, Sept. 7-9, 1915.