130 LEAKAGE FROM STREET MAINS 



These tests of Mr. Brackett's have been often used as a basis 

 for estimating the probable leakage from cast-iron pipe systems, 

 as, for example, the table in the American Civil Engineers' 

 Pocket Book, Third Edition, p. 956. 



Mr. H. F. Dunham stated, in commenting on Mr. Brackett's 

 data: 



The speaker found in making tests of entire pipe systems where the mains 

 ranged from 16 to 4 inches in diameter, and a pressure of 50 pounds per 

 square inch was maintained, that the quantity of water escaping did not 

 exceed i gallon per minute per mile of street main. These tests were 

 made when the pipe systems were new and before the introduction of service 

 pipes. They did not include such breaks as would immediately appear on 

 the surface, but did include some leaks from porous castings, faulty 

 joints and imperfectly packed gates that were noticed subsequently. The 

 quantity was determined by plunger displacement. Trans. Am. Soc. 

 C.E., Vol. XXXIV, p. 204, 1895. 



One gallon per minute amounts to 1440 gallons per twenty- 

 four hours. Assuming that the average size of pipe is between 

 6 and 8 inches, this indicates only about half of the rate of 

 leakage shown by Mr. Brackett's test. 



A statement in regard to leakage from pipe mains by Emil 

 Kuichling, for many years Engineer of the Rochester Water 

 Works, and an engineer with unusual powers of observation, 

 has been widely quoted. 



It should be understood that the leakage here referred to is limited to 

 that which does not show on the surface of the ground, and the individual 

 components of which cannot be detected by the most careful inspection. 

 Loss by wilful waste on the part of consumers and by breakage of pipes is 

 distinctly excluded from present consideration, since the former may be 

 corrected by the use of meters, and the escape of a comparatively large 

 quantity of water at a single point generally renders itself manifest after 

 a short time. The inquiry may therefore be restricted to the loss due to 

 the sweating or slight dripping of the pipe joints, valves of fire hydrants, 

 stuffing-boxes of stop-valves, badly ground taps and curb cocks, and defec- 

 tive joints in service pipes. 



From close observation of thousands of water pipe joints and fixtures 

 in various localities, both when first laid and after having been in use for 

 years, the author has reached the conclusion that a discharge of one drop 

 per second from each joint, five drops from each hydrant and stop-valve, 



