Plate XXXII, Fig. 2.—View of Santa Catarina River Flowing Through Low-Lying Streets, 8 Days After the Flood.
On reaching the city, the flood of August, 1909, swept away two streets adjoining the river. These streets had been built on made ground, in what was originally the river-bed. The sewers and water mains laid in them were destroyed entirely, and some 460 ft. of the 24-in. cast-iron pipe, buried under the river-bed at a depth of 8 ft., were carried away. In relaying this portion of the main, and for protecting the remainder of it across the river, it is now proposed to encase it in a solid rubble concrete block, 8 ft. square, which will impart weight and stability against the scouring effect of floods.
The South Reservoir is circular in shape, with an interior diameter of 165.68 ft. at the top, and is partly excavated in the ground and partly completed by an embankment of vast proportions ([Fig. 10]). Right after the flood of August, 1909, a wet spot appeared on the northeastern toe of the embankment, and it was supposed for some time that it was the effect of the saturation produced by the preceding rains, but, as it persisted for several months, it was obvious that its origin was in the interior of the reservoir, which was emptied when the writer took charge of the work. The first inspection revealed a horizontal crack in the concrete lining, about 310 ft. long and extending about 153° around the circumference on the north side. Throughout its length it coincided with the line of cut and fill. Vertical cracks, coinciding with the panel points in the lining, had also developed, and extended from the main horizontal crack to the roof. The circumstances originating this development can be conjectured by considering the position of the main crack, its characteristic features, and the conditions that preceded its formation. The coincidence of the crack with the joint of cut and fill, points to this line as a source of danger. An examination showed, besides, that the fracture was clean and sharp, ranging in thickness from a hair line at the ends to 3⁄16 in. at the center, and that its upper border projected over the lower one perceptibly, a proof that horizontal motion had taken place. The vertical cracks were a secondary effect, the consequence of the displacement immediately after it was scoured. A fracture was discovered in the floor of the reservoir. It started at the center and branched out into two diverging lines in a radial direction.
The circumstance of two abnormal rainfalls, giving 35 in. in 9 days, the precipitation being concentrated in two periods, not far
apart, of 42 hours and 98 hours, respectively ([Fig. 4]), together with lack of provision for shedding the water from the roof of the reservoir and from the surrounding embankment, lead to the inference that the latter became saturated, increasing thereby in weight and decreasing in stability, especially in its steep inner face. A settlement and the consequent horizontal displacement, under these conditions, was natural. The concrete lining, only 16 in. thick at that height, was not sufficient to sustain the resulting strain, and the main fracture developed, permitting the stored-up water to leak into the bank. In time this seepage found its way under the bottom of the reservoir, softening the ground and producing a slight settlement which caused the crack in the floor. Had under-drainage been provided, as at the Obispado Reservoir, the actual conditions would have been noticed earlier. However, as the embankment is of vast proportions, stable in itself to sustain with a large margin of safety the weight of the stored-up water, there was no actual danger of failure, except for the fact that the material forming the structure, on account of its calcareous nature, is dissolved by water. Long exposure to this condition would, in time, open passages in the embankment, and it is certain that there would be cavings in its interior.
The necessary grouting has been done, and provision is being made for water-proofing the interior of the reservoir and shedding the water from the roof and from the embankment, thus relieving the structure of the consequent strain.
Another place in the works where floods have had a damaging effect is the Estanzuela intake basin, which, when the dam was completed, was filled to the overflow level in order to test its water-tightness. As this basin, when cleaned, was found to be slightly fissured on the north side, it was decided to line it with concrete. As shown in [Fig. 8], the lining does not cover its entire area, but only the central portion, leaving a strip on either side without protection. The flood of September, 1910, coming in greater volume than the previous ones of August, 1909, in passing through the narrow gorge at the entrance, undermined the lining in those places where it was not founded on solid rock. Figs. 1, 2, and 3, Plate XXXIII, show some of the damage caused by this flood. The buoyant effect of the water and the impact of large rolling boulders caused fractures all over the surface, and lifted the concrete lining bodily; but the dam proper, being founded on rock bottom, did not suffer any injury. In the future, in order to avoid the seepage of the ordinary supply, alluded to by the author, the water will be carried to the valve-house in an open rubble concrete channel, lined with cement mortar and built high up against the western hillside. The remainder of the basin will be paved with large boulders.
Plate XXXIII, Fig. 1.—Estanzuela Dam: Broken Concrete Basin Lining.
