water has been assumed to stand at elevation 200. If after the dam is built it is found desirable to store to a higher level than 180 feet, it will be practicable by placing gates in the spillway to raise the storage line to 185 feet or possibly higher, these gates, when the reservoir is full, to be opened upon receipt of telegraphic notification of approaching floods. 70. In various reports, notably the report of the 1905 board (par. 31), stress is laid on the fact that the dip of strata at the lower site is downstream. The stata at the upper site also dip downstream, but, as stated above, there is no danger of sliding. The board believes, however, that on account of the dip of the strata it is advisable at either site to design the dam against possible upthrust due to water gaining entry under the dam. It has been assumed, therefore, that under the heel of the dam such upward pressure will be that due to a head of 200 feet above stream level and that it will diminish uniformly to nothing at the toe. 71. For detailed account of the processes followed in designing the dam, see Appendix D. The dam proposed for the upper site is shown in plan on map No. 4 and in section on figure 3, Appendix D. The photographs herewith show the topography of the canyon and of the sites discussed in the report. (Pls. 2 to 11.) 72. In estimating quantities in the dams at the two sites, it is assumed that below stream level the bedrock is removed to the average depth of 3 feet vertically below the contours indicated by the borings, and that above stream level the excavation at the lower site is to the depth of 7 feet vertically into the rock and at the upper site 3 feet. The reason for the difference in the assumed depths of excavation is that the upper rock strata are exposed to view at the upper site and are known to be of sound rock, while the upper strata at the lower site are covered by a talus of bowlders, preventing examination, but with indications that they are of softer rock (indicated by the easterly borings on lines A and X, and the westerly boring on line X). 73. At the lower site the depth to bedrock is such that the maximum width of the dam measured up and down stream will be about 215 feet. In order to be entirely clear of the fault, the heel, as shown in map No. 4, would have to reach a point about 20 feet upstream from line X. In this position, the contents of the dam would be 236,880 cubic yards, of which 32,011 would be below stream level and 204,869 above. 74. At the upper site, considering the depth to bedrock and the character and configuration of the abutments, the best location for a dam of the height under consideration is as shown on map No. 4, the dam being tangent to line G. This dam would have a volume of t 22,565 cubic yards below the datum plane and 238,781 cubic yards. above, or a total of 261,346 cubic yards. 75. The board believes that a masonry dam of reasonable height could be safely built at either site, yet, notwithstanding that the yardage at the lower site is about a tenth less, the board decidedly prefers and recommends the upper site. At the upper site the depth to bedrock is much less, which, in connection with the somewhat greater width of canyon at stream level, would make the handling of the river during construction less difficult. The distance of the dam from the nearest fault is several times greater than in the case of the lower dam. The rock is quartzite or quartzitic sandstone, while at the lower site it is limestone and contains warm springs, indicating the possibility of cavities. The abutments above stream level are wholly in sight and are sound, while at the lower site the lower part of the right bank abutment is somewhat shattered and may possibly be detached from the mass of the hillside. The indications are that branch of the fault below it runs obliquely upstream approximately parallel to the bluff line and some 40 or 50 feet behind it, and that one or more fractures extend from the face of the cliff back to this branch. Even the advantage of the lower site in respect to the quantity of material in the dam might prove illusory when the preparation of the abutment on the right bank was undertaken, for the removal of a considerable quantity of rock might be found necessary there. Finally the upper site is more accessible for construction purposes than is the lower, and is nearer to the area in which dredging for desilting will be required. 76. Elements of cost. It is assumed that the dam is to be of cyclopean masonry, consisting of random stone of variable sizes embedded in concrete; that 40 per cent of the mass will be random stone and 60 per cent concrete; that concrete will be of 1 part Portland cement, 24 parts sand, and 5 parts broken stone; and that it will be deposited against forms without the use of any hewn stone to make the walls. At the present time Portland cement would cost, delivered at the dam site, about $3.50 per barrel. It is not overlooked that important structures are being built of sand cement. But with cement at present prices or less, cheap power for grinding not available, and cost of plant high for the amount of cement required, there is little inducement to depart from the established practice of using ordinary cement. However, when the time arrives for construction the way may seem clear to use sand cement, for which quartzite is conveniently available; or possibly a mixture of cement and tufa, of which there is an accessible deposit some 6 or 8 miles away. Sand for concrete may be obtained in the river bed at the site, though some, if not all of it, would have to be washed. It may also be made from quartzite crusher screenings. The choice will probably depend on cost of power for crushing, on amount of washing required for the river sand, and on supply of water, for the river is sometimes dry. 77. The broken stone and the random stone would be taken from the location where quarrying would provide a spillway past the easterly end of the dam. The stone, being quartzite, is of good quality for the work. At the high elevation, above 180 feet, the exposed rock is full of small fractures; and quarrying for the random stones may give more small stone than is required for concrete and for intermixing with the larger stones in the construction, thus involving some waste. 78. For hauling material and supplies from the railroad, 7 or 8 miles distant, a wagon road will have to be built on the northerly and westerly side of the reservoir, and as part of it will require heavy earthwork, rockwork, and some bridging, it will be costly. 1 See Plates 4 and 11 Diversion of river.. Excavation of foundation, 26,000 cubic yards, increased to 52,000 cubic yards in anticipation of duplication on account of possible floods, at $1 per cubic yard, including pumping.. $52,000 70,000 Camp, including water supply.. 50,000 45,000 200,000 Road building.. Plant...... Towers, conduits, gates and valves, including power plant for their operation. 300, 000 Total...... 717,000 Seven hundred and seventeen thousand dollars amounts to about $2.726 per cubic yard of masonry in the proposed dam and spillway weir, increasing the cost to $8 per cubic yard. The cost of the dam, not including engineering and contingencies, will then be: 263,000 cubic yards (which includes spillway weirs), at $8, $2,104,000. PROPERTIES OF RESERVOIR SITE. 80. The table below, furnished by the United States reclamation office at Phoenix, and used by the 1905 board in preparing the diagram which accompanied its report, shows the properties of the reservoir site for dams of various heights at the lower (1899) dam site. Capacities are not reduced materially by assuming the dam to be at the upper site. 81. The net loss by evaporation from the reservoir as arrived at in Appendix E will be taken at 60 inches per year. SEEPAGE FROM RESERVOIR. 82. There is nothing in the appearance of any part of the area to be covered by the reservoir to indicate that any considerable loss may be expected from seepage. Accordingly no allowance for seepage from the reservoir is made. Neither is excess storage in the ground counted upon, though this will make a certain quantity of water available at low stages when most needed. WATER SUPPLY. (Appendix F.) 83. The following table shows the run-off of the Gila at San Carlos for all years for which there is direct or indirect information. The derivation of the table is explained in Appendix F. 84. While the run-off data are not as complete nor probably as accurate as could be desired, they indicate a variation in yearly discharge of from about 100,000 acre-feet to more than 10 times that amount, and a probable mean annual run-off of about 346,000 acre-feet. 85. With a dam 180 feet high, giving a reservoir of about 709,000 acre-feet capacity, a portion of the run-off of 1905 would, of course, have passed over the spillway. The same would probably be true of at least an occasional run-off of 600,000 or even 500,000 acre-feet. Allowing for this, the average amount that could be stored yearly by a dam 180 feet high would probably not exceed about 310,000 acre-feet. On the other hand, not all the water which passes over the spillway in excess of the usual draft would necessarily be wasted, since a portion of it at least would be taken into the main canal. The main canal will have a capacity greater than that required to carry the amount usually applied to irrigation, and the excess water would be used for such purposes as seem best at the time. 86. The most striking fact disclosed by the table is not the great difference between the maximum and minimum run-offs, but the length of the periods during which the annual run-offs are continuously below the mean. Thus the run-off was below the average each year of the three years 1908-1910 and of the seven years 1898-1904. The need for hold-over storage is apparent. SILT. 87. In the earlier determinations of the amount of silt in the Gila, the method followed was to allow the muddy water to settle for a day, or for several days, in a test tube, and to note the volumetric relation of deposited silt to the water sample, and then to divide the percentage thus obtained by 5 in order to reduce to the percentage of volume which the silt would finally occupy as soil in a reservoir. Though this sample method is useful, in a general way, for making comparisons, and is doubtless correct for some degree of muddiness of water, it does not apply to all the varying degrees of muddiness nor to the records obtained by test tubes of different heights. Columns of silt of different heights have different densities, dependent on the compression due to their own weights. The size of the test tubes used also has some effect on the degree of settlement of the deposit. Mr. Hughes's experiments show that the divisors for reduction of deposit in tube to soil vary from as low as 5 for the highest percentages to perhaps as high as 10 for the small percentages. The density of silt in the test tube depends also on its specific gravity, its fineness, and other physical properties. 88. The most approved method for making silt observations is to take samples of muddy water of given volume or weight, allow the sample to settle, decant the clear water, and evaporate the remainder and then weigh the silt. In passing from weight of silt (as compared with the weight of the sample of water) to the volume which it will occupy in a reservoir, Mr. Hughes's figures, determined from 15 samples, give 70 pounds per cubic foot as the average weight of the dry material in place in the reservoir. Mr. W. W. Follett's use of 53 pounds per cubic foot was based upon a single observation only.1 1 Silt in the Rio Grande, by W. W. Follett. |