Again, unless the rock is exceptionally favorable, or the tunnel deeply buried (i.e., kept some distance away from the foundation of the dam) the process of excavating the tunnel may open seams in the rock between it and the dam that may later become a source of leakage.

Considered solely as a temporary diversion channel it must probably be said of such a tunnel that the capacity is limited, and that it is expensive. Upon most alternate schemes much work might be done previous to the installation of the plant necessary for the rapid and economical construction of the tunnel.

The tunnel must be driven, probably lined, and certainly provided with a screen at the entrance, before much if anything can be done toward diverting the stream. A screen is specified because a tunnel without one might possibly be blocked by drift in the tunnel, which would result in very serious if not fatal expense and delay. Though the screen be readily accessible it will be difficult to keep clear, in proportion to the amount of drift. In fact, should lumbering operations be going on along the river, it might be a practical impossibility to do so.

While conditions may indicate unmistakably that a tunnel is the proper solution, earnest consideration should be given to the alternative of an open channel crowded into the side slope or wall far enough to clear that area of foundation lying below the elevation of the bottom of the channel. This would involve the excavation of a notch in the rock along one side of the river at about water level, one side of the notch being the floor and the other the vertical uphill side of the open channel. Then erect a masonry wall for the other side.

The notch and wall may be amply long enough to connect with feasible temporary dams and still not be as long as a tunnel. There might be many more cu. yd. of excavation, but that part of it across the dam-site would practically all be necessary in any event in preparation of the foundation. Even if the net amount is much greater it would be open cut and much cheaper than tunnel cut. Any possible disturbance of the rock under the foundation of the dam would be entirely eliminated. Any temporary or permanent gate frames or guides could be more readily set and be more readily accessible. In the much less likely event of an obstruction, it would be a simple and easy matter to clear an open channel. Whenever desirable the channel could easily be closed, reduced in size to any dimension for permanent outlet, arched over so that work could proceed above it until later closure, or permanently closed and the stream diverted to a similar higher channel.

Actual Examples with Cost - Type 1 as illustrated by the scheme of diversion employed at the McCalls Ferry dam across the Susquehanna River.

At this point the river is 1/2 mile wide, divided into two channels by a low island, about in the middle, and has 26,766 square miles of watershed. The flow of the river at times is as low as 3000 c.f.s., and the maximum flood is nearly 700,000 c.f.s. A study of the stream records indicated that it would be wise to provide for handling 50,000 c.f.s. without interruption to the work. The dam is about 50 ft. high and has an overflow section for practically its entire length, i.e., 2350 ft.

Cofferdams diverted the entire flow into one of the channels, while in the one unwatered were built the powerhouse foundations and one-half of the dam. While the bottom was all prepared and the foundation masonry all put in up to a certain elevation, the dam was built in alternate sections of 40 ft., leaving an equal number of 40 ft. openings through which to divert the stream later. The work was handled by cranes traveling on a construction bridge just below the dam. In order that no delay might be occasioned by possible floods carrying away the bridge it was built of concrete in the most substantial manner. (See Plate II, Figs. C and D).

This bridge has been criticised because it was so expensive (about $100,000) compared with the figure at which a steel bridge could have been erected. However, the size of the floods encountered during construction amply demonstrated that some kind of a bridge was the best solution of the problem.

The cofferdam was composed of cribs 16 ft. wide by a length up-and downstream one and one-half times the depth of the water. The cribs were spaced 10 ft. apart in the clear and after being weighted the openings were closed by 12 in. square stop timbers. Along the upstream face was placed a double line of vertical 2-in. sheeting, and puddling material above that. The cofferdam thus constructed was remarkably tight, for 1000 ft. of it under 18 ft. head leaked less than 1 c.f.s.

When the first half had been built as above described, i.e., alternately 40 ft. brought up to the top and 40 ft. left open above the foundation, a precisely similar process was gone through with on the second half, diverting the river through the openings left in the first half. When the entire dam was brought to the stage of alternate 40 ft. sections completed and open, the openings were brought up 5 ft. at a time. To accomplish this each opening was provided with a very ingenious shutter, the water was shut off, and 5 ft. of masonry was placed and allowed to harden until the water could be passed over it. Entire completion simply involved repetition of the process. Several considerable floods were passed during the construction period with but slight damage and only the unavoidable loss of time.