46 BULLETIISr 376, U. S. DEPAETMENT OF AGEICULTUEE. 



within the pipe, are to occur. The capacity of tlois pipe was 9 per cent less than the 

 discharge computed by the new formula. From the fact that the inlet is from a 

 reservoir, and because the curves are very gentle, experience indicates that this 

 pipe should have a greater capacity for a given loss of head than the new formula 

 would indicate, although it is true that the interior condition ^.f the pipe was not 

 known. Two air valves at summits and one blow-off at a Ioav point occur within the 

 reach tested. The air valves are boxed to prevent freezing. 



No. 51, Expt. S-3, 144-inch Continous-StaTe Douglas Fir Power Pipe 

 Line,! Salmon River Power Co., New York.— About 4 miles from Altmar, N. Y., 

 on the bank of the Salmon River, is located an hydroelectric plant, constructed in 

 1913 and 1914 to carry a portion of the load formerly served by one of the big plants 

 at Niagara Falls. Stillwater Reservoir is formed by a dam across the main channel 

 of Salmon River about 2 miles above the power plant. A tunnel 600 feet long con- 

 veys water from the reservoir to the upper end of a continuous-stave Douglas fir pipe 

 line 144 inches (12 feet) in diameter. (PI. V, fig. 1.) At the end of 3,450 feet a taper 

 transition section about 50 feet long leads into a similar pipe 132 inches (11 feet) in 

 diameter. Tests were made on the 12-foot pipe. The portion of the pipe tested is 

 without vertical curves, being laid on an even gradient Practically the lower third 

 of the pipe is buried. No leaks worthy of notice occurred throughout the pipe. 

 The line had been in operation but a few months, and since the velocities were high 

 (up to more than 8 feet per second), it probably was in perfect condition on the inside, 

 although it was not feasible to ascertain this. The staves are 4 inches thick. Taps 

 for the nipples were made by a Yg--inch wood bit until the tip of the bit punctmred the 

 inside surface of the pipe. The nominal area of the pipe was accepted as its true area. 

 After making .these tests the writer made careful measurements of a still larger pipe 

 built by the same company and found the true area extremely close to nominal area. 

 The discharge of the pipe, from which the velocity within the pipe was obtained, 

 was determined in the following manner: As~ shown in Plate V, figure 2, after the 

 water passes through the turbines it falls over a submerged weir into a tail-race 

 channel which in turn discharges into Salmon River about a quarter of a mile 

 below the power house. A good meter rating could be obtained, as the mean velocity 

 for the greatest discharge was but 3.25 feet per second. This velocity did not cause 

 a tiu'bulent condition in the channel, since the latter had a hard, fiat rock bottom. 

 The form of the weir and the conditions of velocity of approach are such that the 

 writer did not feel justified in accepting the discharge as computed from any known 

 weir formula. The velocity of approach, in particular, is an uncertain quantity, 

 since the bottom of the charmel slopes up from 20 feet deep at the power house to a 

 mean of but 1.234 feet deep immediately above the weir crest, within a horizontal 

 distance of 220 feet. The weir is 79.6 feet long with end contractions approximately 

 suppressed. It is a concrete wall 18 inches thick, rounded over on top. It is not 

 designed as a measuring weir but for the sole purpose of drowning the draft tubes 

 for all discharges. This weir was caUbrated by making four careful current-meter 

 measurements with as many discharges from a bridge across the tailrace below the 

 weir; and meanwhile reading a hook gauge in a stilling box 8 feet above the weir and 

 a tape gauge 3 feet below the weir. The latter gauge reading had no bearing on the 

 caUbration but was taken for the purpose of secm-ing more information concerning 

 submerged weirs. The results of these measurements follow. The elevations are 

 based on a bench mark with an assumed elevation of 10.000 feet. 



1 Eng. Kec, vol. 69, No. 24, June 13, 1914, p. 671. 



