THE FLOW OF WATER IX WOOD-STAVE PIPE. 
47 
Table 4. — Simultaneous discharge, elevation of water surface above weir, elevation of water 
surface below iveir, and head on weir. 
No. 
Discharge. 
Elevation 
above weir. 
Elevation 
below weir. 
Head on 
weir. 
1 
Second-feet. 
106.6 
446.1 
724.9 
749.6 
Feet. 
8.752 
9.539 
10.123 
10. 172 
Feet. 
7.77 
(not taken) 
9.62 
9.62 
Feet. 
0.515 
2 
1.302 
3 
1.888 
4 
1.937 
The mean elevation of the weir crest, 8.237 feet, was based on readings with level 
and rod taken every 5 feet throughout its length. During measurement No. 1 the 
hook gauge remained constant. During No. 2 water rose 0.142 feet on the weir. 
During No. 3 water fell 0.020 feet on the weir. During No. 4 water fell 0.049 feet 
on the weir. The mean gauge reading was accepted where fluctuation occurred. 
Current-meter measurements were made by the two and eight-tenths depth 
method. As the load carried (and consequently the discharge of water at a 
power house) varies throughout the day, and since the discharge is controlled 
by the load (by means of governors), the following method of testing the^l2-foot 
pipe for loss of head was adopted: The mercury manometers and the hook and tape 
gauges were read continuously throughout the morning and afternoon. A synchronous 
profile was then platted showing all gauge readings. From this profile periods of 
comparatively uniform flow were chosen and each of these periods was designated as 
an observation. These would necessarily vary in length of time. From the calibra- 
tion curve of the weir the discharge for each reading of the hook gauge was taken 
and the mean of these discharges was assumed as the discharge which held throughout 
the observation. The capacity of this pipe was 2.5 per cent less than that computed 
by the new formula. Since the pipe was new, joints smooth, and the curvature 
gentle, the writer wouM estimate the capacity of this pipe to be greater than that 
computed by the new formula. Tests by all experimenters show similar cases where 
the observations indicate far different results than the conditions appear to warrant. 
No. 52, Expt. S-14, 162-incli Continuous- Stave Douglas Fir Power Line, 
Northwestern Electric Co., Condit Plant x on White Salmon River, Wash- 
ington. — About 2 miles above the mouth of White Salmon River is located the 
Condit Plant of the Northwestern Electric Co. Some 6,000 feet upstream a high 
diversion dam raises the water above the intake to the supply pipe line. This, said to 
be the largest wood-stave pipe in the world, 162 inches or 13J feet in diameter, is used 
to convey the waters of White Salmon River from the diversion dam to the surge tank, 
a distance of 1 mile. Within the surge tank is a structure that divides the water from 
the 13^-foot pipe between two 9-foot pipes with very little loss of head. Each of the 
9-foot pipes serves 1 electrical unit in the power house. (PI. XI, fig. 1.) About mid- 
way of the large pipe, upon which tests were made, is a bend of 83° with a radius of but 
40 feet (less than 3 diameters) . In the opinion of the writer, such a bend would cause 
an appreciable loss of head independent of the friction loss, and for this reason a 
reach of pipe was chosen between this bend and the surge tank. Mercury manometers 
were used for both gauges, the equivalent water column being just too high to be 
feasible. (PI. V, fig. 3.) Gauge No. 1 was located on the zone of neutral velocities 
209.9 feet from the bend. Gauge No. 2 was located 2,378.9 feet from gauge No. 1 and 
about 40 feet above the dividing tongue in the surge tank. During all of the runs the 
load carried by the unit served by the right-hand 9-foot pipe was held constant, all the 
fluctuation being thrown to the other 9-foot pipe. The time necessary for fluroescein 
to travel from gauge No. 2 through the constant-velocity pipe was determined by 
i Eng. Rec., Oct. 11, 1913; Eng. News., vol. 70, No. 15, p. 685. 
