Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec. 035 0.32 043 Base Line The heavy line represents the mean monthly water supply in acre-ft. at San Carlos for 19 years (seasonal year Sept. 1,1889 to Sept. 1,1890 and calendar years 1895 to 1912 inclusive). The light line represents the mean monthly rainfall in inches at 14 stations over the water shed including all observations prior to 1913. Diagram No.8. San Carlos Irrigation Project. San Carlos and Cliff, and Alma plus 20 per cent. W. S. P. No. 175, pp. 162 and 170. 1902 .....do. 1903 .....do. 1904 .do. 1905 1906 1907 1908 1909 Supplied from rainfall and mean of same months Redrock and Alma plus 20 per cent.. 1910 Redrock and Clifton plus 10 per cent (Clifton 1911 W. S. P. No. 211, pp. 123 and 128. W. S. P. No. 249, pp. 176, 178, 179, 180. W. S. P. No. 269, pp. 221 and 224. W. S. P. No. 289, pp. 202-207, 208, and blue-print 1 Kelvin in 1911-426,220; in 1912-450,691. D. E. HUGHES. C. H. MCKINSTRY. APPENDIX G. AMOUNT OF SILT THAT WOULD BE DEPOSITED IN A RESERVOIR AT SAN CARLOS. 1. The silt problem, being next in importance to the sufficiency of the water supply and the suitability of the dam site, requires the most careful analysis of all available data. 2. The board, at its first meeting, decided not to make observations of the river's flow, knowing that a few months' work could add but little to the more extensive data already available. But a reading of the different silt determinations here and elsewhere, and the conclusions based thereon, suggested the need for further observation and experimentation to make more useful the varying data obtained under different conditions, recorded in different terms, and reduced under different hypotheses. For instance, volumteric percentages have, in different times and places, been determined in tubes. of different diameters and of different lengths, and after settlement for different periods, such as: "Till water is clear," "Till settlement ceases," "Several days," "One day," "Seven days," and "One month." And different factors have been used to reduce such volumetric percentages to the volume of soil such mud would make. Even words used by one have been misunderstood by another, thus: "Solids," as used by one to represent resulting earth or soil with its voids, as is clearly indicated elsewhere in his report, is assumed later by another to mean an absolute solid without pores or voids. Again, in reducing percentages by weight to volume of earth the sediment would finally make, the earth has been assumed to weigh all the way from 50 to 120 pounds per cubic foot in place after being dried, notably, 53 pounds, 62 pounds, 85 pounds, and 100 pounds. Not all the above variations apply to the data on the Gila River, but they are to be borne in mind when comparing various findings on different rivers. 3. The observers in the past would now, at the present stage of the evolution in silt determination, doubtless agree that volumetric percentages indicated depend on diameter of tube, for the smaller tubes, by the relatively greater influence of the friction or adhesion of their walls lessen the progressive daily settlement of the mud contained. and on length of tube used, for the higher column of mud must settle relatively faster through having its lower portion more compressed by the weight of the mud above. Also that in like tubes, the character of the silt, including its specific gravity, remaining the same, the higher the column of mud the greater the reduction from day to day in the indicated percentage. 4. They would agree, too, that if any divisor, as 5, will correctly reduce a column of mud to its final volume in reservoir soil, a higher column of mud would, on account of its greater initial compression. require a lower divisor, while lower columns would need divisors of 6 or 7 or more. Also that, for like reason, there can be no fixed relation between the volumetric percentage and the percentage by weight, and for the further reason that percentage by weight depends also on the specific gravity of the silt material, which may vary from perhaps as low as 2.1 to as high as 2.7. 5. And there would be agreement that weight of soil in place after being dried is proportional to the specific gravity of its material, and also dependent upon percentage of voids; and that after excess water has once escaped through drainage or evaporation the percentage of voids can hardly exceed 56 per cent, nor in soil not subjected to great compression fall short of 45 per cent, except in case of admixture with coarse sand or gravel; and that the variation between the 56 per cent and the 45 per cent depends largely on circumstances of deposition-whether the silt particles are graded as to fineness to be deposited in different localities by a free river, which would make the voids greatest and the soil lightest; or whether all grades of silt are constrained to be intermixed in a deposit behind a dam, which would result in somewhat denser, heavier ground. A cubic foot of earth in place with 56 per cent voids must weigh in pounds, when dried, 27.5 times the specific gravity of its material. With voids of 45 per cent the multiplier becomes 34.3. 6. Fine silt, before losing its excess of water, may have voids as high as 60 or 70 per cent or even higher. Such silt deposited beyond the reach of sand, in harbors where it can not be drained, remains so light and soft that it is hard to tell by sounding just where its surface lies, and vessels plow through a foot or more of it without material hindrance. But such condition can not long apply to a reservoir occasionally emptied, for after the excess water once departs the superior gravity of the earthy matter prevents readmission of water beyond the capacity of the voids. Dry clay swells a little on being wetted, but repeated drowning and draining reduces its volume. 7. Soil below the surface is somewhat compressed by the weight of the soil above, but not to any great degree under the moderate load that would here obtain. In fact, there can not be within the depth and life of any artificial reservoir a degree of solidifying approaching the transition from silt to shale or sand to sandstone. In any event, nothing approaching the weight of sandstone could be used as a factor in reducing silt, for the increase in weight of sandstone comes not so much from compression of sand as from the infiltration of heavy cementing material displacing water in the voids of the sand bed. 8. Though 55 is the highest percentage of voids found in practice when depositing angular fragments of broken stone of equal screen size, surface earth, no longer subject to overflow, may be so burrowed with vegetable roots and animal life as to make its voids 60 per cent or even more, when the organic matter is expelled. But such conditions can not long survive subsequent depositions, if any, over the same area, for the more porous and permeable ground becomes to a degree a filter bed for the muddy waters until the voids are again reduced. 9. On reflection all would also agree that though the lower strata of silt are compressed by the weight of the upper strata, none of them are at all compressed by the weight of superincumbent reservoir water, for the water maintains a balanced hydrostatic pressure throughout the voids. The ocean bed beyond the reach of sand, when not of rock, is found to be the softest ooze, though depth be miles and years be countless. Only through occasional exposure to evaporation or drainage to remove excess water can the upper strata of deposited silt be reduced to the consistency and weight of soil, for the weights of silt particles as compared with their surfaces are too small to overcome the adhesion of water and squeeze it out until they come to a bearing on each other. 10. During August, September, and October, 1913, several experiments were made with the silt deposits and with the silty waters of the Gila, to verify and illustrate the foregoing theoretical deductions, and to add a little to the silt data on hand. 11. Fifteen samples of soil deposited from Gila River waters in different localities were accurately determined as to volume occupied in place, and as to specific gravity and weight after being completely dried. 12. Samples Nos. 1 to 4 were taken August 12 from the bed of Picacho Reservoir near its dam at distances of 1, two-thirds, onethird, and 0 mile from its outlet. This reservoir, situated between Florence and Casa Grande, has been in use for 25 years storing surplus Gila River waters from the Florence Canal. Last April some water flowed through it over the localities from which the samples were afterwards taken, but on August 12 it was empty, except for a few gallons at the outlet, where sample No. 4 was obtained. 13. Samples Nos. 5 to 8 were taken August 19 in the silt deposits in the San Carlos Reservoir site along or near the wagon road leading from San Carlos to the dam site. Samples Nos. 9 to 12 were taken October 9, near lower end of San Carlos Reservoir site, in very old river deposits, where a comparatively recent narrow wash 15 feet deep extends from the river a few hundred feet along a cattle trail. Samples 13 to 14 were taken October 13 in old river deposits on left bank of river near road leading from San Carlos to Copper Reef; and sample 15 was taken between camp and dam site. 14. Special care was taken in measuring samples in place. After top earth was removed to get below shrinkage cracks, if any, an annular excavation was carried deeper around undisturbed earth left in the form of a frustum of a cone with level top. On this frustum a sheet-iron pipe was pressed or gently driven to cut a cylindrical sample a few inches long, the earth outside the pipe being trimmed down from time to time when necessary to facilitate the descent. Light tapping on a board held on the pipe was more satisfactory than pressure in overcoming friction without distorting any of the material inside. After undercutting and trimming the bottom off flush with the pipe, there resulted cylindrical samples of known diameter and length of undisturbed earth whose volume in place was thus accurately known. The samples were dried in a cook-stove oven for six hours or more until periodical weighings ceased to show loss. The weights, specific gravities, etc., are in the table following. |