Part 4

_Excavation and Embankment._--The heavy slope of the ground at the selected site made the circular form the most desirable. On the low side the ground was excavated about 2 m. below the original ground line, while the excavation at the upper part of the slope was about 12 m. deep. The excavated material consisted chiefly of sillar and limestone conglomerate, which when broken up forms a calcareous clay of an excellent character for the formation of embankments, when proper care is taken. The dimensions fixed for the internal diameter of the finished concrete work of the reservoir were: 81 m. (265.68 ft.) at the top, and a depth of water of 9 m., with sides sloping 55 in 100.

Fig. 10 is a plan of the reservoir, with a cross-section of the excavation and embankment. On the lower side the original ground line was cut down in steps, and all loose earth, roots, etc., were carefully removed. The floor of the reservoir was chiefly sillar conglomerate, a hard material that required a considerable amount of blasting for its removal. The embankments were formed in 10-cm. layers of sillar and conglomerate broken into small fragments and then rolled with 3-ton sectional rollers drawn by teams of 4 and 6 mules, which assisted in disintegrating the mass thoroughly, and produced by constant wetting a homogeneous and compact clay. The excavation and embankment were left so that 15 cm. of trimming could be done at a later date, immediately prior to the lining of the reservoir. The excavated material amounted to about 34,000 cu. m., and, of this quantity, 31,500 cu. m. were used to form the embankment; the remainder was taken to a spoil bank immediately adjoining, the black earth stripping being separated and reserved for covering the reservoir, etc. The contract prices for the excavated material placed in the embankment were:

Pesos per cubic meter

Class 1.--Material which could be removed by plows and scrapers 0.60 Class 2.--This consisted chiefly of "sillar" 1.09 Class 3.--Limestone conglomerate (requiring blasting) 1.65

The prices (for the same classification) for material taken to the spoil bank, were 0.40, 0.80, and 1.40 pesos, respectively. Of the material taken out, 15% came under No. 1 classification, 80% under No. 2, and 5% under No. 3.

The excavation was begun at the end of May, 1907, and completed in January, 1908, by Scott and Lee, the contractors. The embankments were then allowed to stand until the beginning of July, 1908, to permit the whole to become thoroughly settled and consolidated prior to beginning the lining. In July the work of trimming the embankments and excavating for the foundations of the reservoir columns was commenced, under the Company's own administration, which completed the entire work.

_Concrete Lining and Roof._--The general arrangement and details of the side-walls, columns, and roof are shown on Plates VI, VII, VIII and IX. The principal feature consists in dividing the reservoir into radial sections and supporting the roof on 135 primary and 670 secondary beams, from 135 columns, spaced as follows:

Outer ring, at 34.25 m. from center 40 columns. 2d " " 27.88 " " 40 " 3d " " 21.51 " " 20 " 4th " " 15.41 " " 20 " 5th " " 8.77 " " 10 " 6th " " 2.40 " " 5 " --- Total 135 columns.

The inner bottom diameter of the reservoir is 70.32 m. (230.64 ft.); the upper inside diameter is 81 m. (265.68 ft.); the water depth at the overflow level is 9 m. (29-1/2 ft).

The roof was designed to carry a dead load (the earth cover) of 150 lb. per sq. ft., and a live load of 100 lb. The maximum compressive fiber stress in the concrete was assumed at 550 lb. per sq. in. for the beams, and at 350 lb. for the columns, a low figure, because of their eccentric loading. The tensile strength of the steel was taken at 14,500 and 16,000 lb. per sq. in. The twisted steel used for the column reinforcement was made at the local steel plant, but for the beams, etc., a twisted lug bar, of higher quality and greater permissible tensile stress, was used. The total quantity of steel used was 178 tons. It was calculated that the load on the column foundations would not exceed 1-1/4 tons per sq. ft. With the exception of the side-wall and floor, all the concrete was reinforced with steel, of the sizes and spacing shown on Plate VI.

_General Construction and Erection Scheme._--The question of ordinary forms, requiring very heavy timber work, was a serious one, as suitable lumber is very expensive in Mexico; and the necessity of finishing this reservoir before the end of the first term allowed under the concession, which expired on December 31st, 1908, led to the adoption of what the writer believes is an original scheme for so large a structure. This scheme was to cast the columns in short sections, mould the radial and secondary beams as separate members, and then place them in position with derricks. At the same time, in the case of the beams, it was important not to sacrifice either the benefit of that part of the slab which is ordinarily assumed to act as a part of the beam, or the additional strength due to continuity; and, in case of the columns, the strength due to the reinforcement extending from the foundation to the beams.

The T-beam section was secured by notching the tops of the moulded members, with notches 10 cm. deep, throughout the lengths of the beams, as shown on Plate VI. A computation of the maximum flange increment shows that these notches are sufficient to transfer the flange stresses to the stem, but, for additional security, flat steel bars were bent to a Z-shape and embedded in the top of the beam, about 60 cm. apart. Continuity in the beams was secured by carrying the steel to the tops of the beams over all supports, and, after erection, concreting them into the roof slab. The secondary beams, after casting, were dropped into recesses left in the radial beams for the purpose.

_Concreting, Mixing, etc._--The radial beams and column sections were cast as nearly as possible under their ultimate positions; the secondary beams were cast outside and immediately adjoining the reservoir.

The rock and sand was brought from the Company's crushing plant, in 3-cu. yd., side-dump cars, running on a 30-in. track by gravity a distance of 1 km., the last 150 m. requiring hauling with 6 mules. The cars returned all the way to the crusher by gravity. These cars dumped the material into bins on the high ground above the reservoir; from there it was hoppered into cars which carried to the mixer all the material for one batch of concrete. Two No. 1 Smith mixers were used, and from 25 to 30 batches per hour could be handled in each machine.

The concrete was transported from the mixers to place in 1/2-cu. yd., 18-in. gauge, swivel, steel dump-cars pushed by two men. All the concrete used in the bottom of the reservoir, for the main beams, columns, and floor, amounting to about 2,460 cu. m., was dumped through a chute into smaller cars. The chute had so many baffle-plates and bolts that it resembled a gravity mixer, but, although it was 12 m. long, it effectively prevented the separation of the materials.

_Concrete Placing and Moulding._--The square foundations for the columns were deposited _in situ_, a recess being left for the reception of the pedestals, which were moulded in place afterward. The capitals and pedestals were cast in one piece, and the columns in 1.21-m. (48-in.) sections, eight 5-cm. holes being left in them by using wrought-iron pipes, held in place by templates and removed when the castings were about 3 hours old. The columns were erected by threading them on the 15.8-mm. (5/8-in.) reinforcing rods, which extended from the pedestals up through the capitals. The rods were in two lengths, arranged to lap alternately at one-fourth, the center, and three-fourths of the height of the columns. In erection, a light timber frame was used in conjunction with the derrick, and, as the columns were placed, the reinforcing steel was grouted solid with 1:2 cement mortar.

All the erection was done with a combined stiff-leg or guy derrick, having an 80-ft. boom and a 50-ft. mast, and fitted with a 30-h.p. Lambert hoisting engine. The derrick was erected seven times at the circumference, and its final position was on top of the center columns. The moving of the derrick a distance of about 45 m. and its subsequent erection occupied usually about 48 hours. The erection work was carried on continuously, day and night, the placing of the whole of the radial and secondary beams and columns occupying 2-1/2 months.

_Forms._--As the erection scheme was designed to reduce the cost of forms, economical construction was of considerable importance. The wall was formed in 40 panels, about 6 m. wide and 11.27 m. high. The chief object in arranging them in this manner was to permit an expansion joint, 30 cm. wide, at each panel; this joint was not filled until after the completion of the roof, when the temperature inside the reservoir was uniform and not subjected to such great fluctuations as if exposed alternately to the hot sun and comparatively cool nights. The range of temperature during the construction period sometimes amounted to 80° Fahr. in 24 hours.

The expansion joints were left to the last, when a uniform temperature of about 70° inside permitted the filling of the joints, thus avoiding all trouble from expansion cracks.

The forms are shown in detail on Plate VII. They consisted of shutters stiffened with four trapezoidal trusses. The bottom posts of the trusses were fixed in holes formed in the foundation block; they were propped back from the embankment at the top, and secured to anchorages by iron rods.

Six sets of these forms were used to construct the whole wall. The concrete was placed in position through stove-pipe chutes, 20 cm. in diameter, in continuous layers, the workmen treading and spading it well as it was deposited. The forms were allowed to remain 4 or 5 days, and were then struck and removed to another section. The pedestals and capital forms were of lumber, and five of each were used to cast the total number required. In the column sections the outer steel forms used in the manufacture of the Estanzuela pipes were adapted for this purpose. The radial beam forms, shown on Plate VII, were arranged with internal wedge-shaped blocks to mould accurately the recess for the secondary beams. The bottom forms were left attached to the beams for 28 days, but the sides and ends were removed after 24 hours. Eight forms were sufficient for the whole 135 beams.

For the secondary beams, 29 forms were used for the 670 beams, the bottom lumber also being left until they were mature for handling.

By referring to the cross-section of the secondary beam, it will be noticed that it is jug-shaped, shelves being left on either side for the support of the roof forms, which were placed after the secondary beams had been properly grouted to the radial ones. The lagging was laid diagonally, so that the short diameter was slightly greater than the distance between the beams. This greatly facilitated the removal of the lagging, as it was merely necessary to strike the wedge-shaped fillets beneath, and turn them clockwise, after tearing out the end lagging.

The writer believes that the adoption of forms of this type, rather than the ordinary kind, led to a saving of lumber of about 400,000 ft. b. m. During the erection and placing of the concrete, all the joining surfaces were carefully picked and cleaned, particular care being taken at the junction of the secondary with the radial beams, and the upper surfaces of all beams before laying the roof slab.

After the greater part of the roof was completed, the floor was laid in those sections where it was protected from the sun's rays. The concrete was placed in two 15-cm. thicknesses, and the work was carried on night and day, without any joints. The laying of the floor occupied 8 days, or an average of nearly 100 cu. m. daily.

_Proportions of Concrete._--All the concrete work was brought to a smooth face by careful spading, no plastering being used throughout the reservoir, except in the superstructures. The work was kept well watered in every case for about 15 days. The whole of the concrete work in connection with the reservoir was completed in 5-1/2 months. The concrete for the columns and foundations was a 1:3:5 mixture, the aggregate consisting of equal parts of 19-mm. (3/4-in.) and 38-mm. (1-1/2-in.) crushed stone. The remainder of the concrete, except that for the roof, was a 1:2:4 mixture, the aggregate also consisting of equal parts of 19-and 38-mm. stone. With the exception of a short length of the side-walls, the sand used was that manufactured by the Company. When the crushing plant was unable to produce all the sand required, the Hornos sand (see Table 3) was used in the side-walls in equal proportions with the crusher sand.

_Reservoir Outlet and Entrance Tower._--The outlet, 61 cm. (24 in.) in diameter, leads from a well in the center of the reservoir and passes under the floor and embankment to an outside valve-pit, 89 m. from the center. This pipe was laid in a trench in a solid cutting before the construction of the embankment, and was encased in 1:4:8 concrete. Where it passes under the embankment a 1:2:4 concrete cut-off wall, 3.6 m. wide, 2.5 m. high, and 1 m. thick, was placed across it at right angles. The cast-iron pipe is curved upward in the central well, and has a bellmouth on which rests a movable circular copper screen.

Above the outlet well, and on the roof of the reservoir, there is a central tower, giving access to the interior by a steel stairway. This tower also serves as a main ventilating shaft, and in it are arranged the guide-screens and gearing for raising them for cleaning purposes. In addition to the ventilation provided in the tower, 20 circular openings, 30 cm. in diameter, are carried through the roof of the reservoir at the circumference and into the parapet walls.

_Inlet Gate-House, etc._--The inlet gate-house is above the reservoir and about 54-1/2 m. from its center. The conduit enters at 589.00 m. above datum, and the gate-house contains the valves for controlling the inlet pipe to the reservoir, the by-pass, overflow, scour-out pipe, and the copper screens. The inlet, which is 45.7 cm. (18 in.) in diameter, is of cast-iron flanged pipes, carried on iron hangers on the side-wall of the reservoir, and, at a point 90 cm. above the floor level, it is turned at right angles to the side-wall and carried on concrete piers to the center of the first row of columns. The end of the pipe is closed by a blank flange, and the water is deflected at right angles through two 30-cm. (12-in.) branches, for the purpose of setting up a slight circular motion as it enters the reservoir.

The valve-pit is clear of the embankment, and in it are brought together the main supply and by-pass pipes on which are placed two 61-cm. (24-in.) sluice-valves; and between them there is a 20-cm. (8-in.) scour-out pipe, for emptying the reservoir into an adjoining arroyo. The arrangement of the valves gives complete control over the contents of the reservoir.

_Venturi Meter-House._--Fig. 11 shows the arrangement of the Venturi meter and its automatic register in a house over the main supply pipe. This house is designed to form a feature of the entrance gateway of the reservoir grounds, which cover an area of 12 hectares.

_General._--The roof of the reservoir has been laid out as a garden, and the embankments are turfed. The intention is to develop the Company's land as a public park, as it commands fine views of the city and the surrounding mountains. An inspector's house has been built, and a private telephone line provides for communication with the Estanzuela intake and also with the general offices in the city.

SAN GERONIMO GRAVITY SUPPLY.

_Provisional Supply._--It has already been stated that the Company began operations at San Geronimo in March, 1906, by sinking a well on the north bank of the Santa Catarina River at San Geronimo. At this point, a little later, a small steam pumping plant, sufficient to handle about 8,000 liters per min., was installed. The lowest depth to which this well was ultimately sunk in water-bearing strata, was 7 m., the normal level of the water during 1906 and 1907 never falling lower than 569 m. above datum. Tests made from time to time during 1907-08, showed that this well was capable of supplying nearly 10,000,000 liters (264,000 gal.) of water daily.

The excellent supply yielded by this well made it desirable to adopt it immediately as a provisional measure, pending the completion of the larger works forming the western source of supply. To utilize the well to its fullest extent, a reinforced concrete reservoir, of 3,000,000 liters capacity, was constructed on the south bank of the river, the top water level being 585 m. above datum, that is, at the same elevation as the proposed reservoir for the Estanzuela supply. The reservoir is 53.80 m. long, 21 m. wide, and has a water depth of 3.25 m. at the overflow level. It is excavated on a steep hill slope, and has an earth embankment on the lower side. The lining is of concrete, 20 cm. thick, and the roof is of reinforced concrete composed of flat arches springing from beams carried on 46 by 35-cm. reinforced columns. There are 68 of these columns, and they are 3 m. apart longitudinally and 5 m. apart transversely. The roof was not constructed until October and November, 1907, and prior to that time the necessity of covering the reservoir was amply demonstrated by the growth, during hot weather, of considerable quantities of green algæ, which had to be skimmed from the surface of the reservoir every few days.

The delivery pipe from the pumping plant was originally of riveted steel and was asphalted. It was 30 cm. in diameter, 2 mm. in thickness, with slip joints, and where it crossed the river it was encased in concrete. This pipe was afterward replaced by a cast-iron pipe of the same diameter. The supply pipe to the city was also of sheet steel, 30 cm. in diameter. For a part of its length it was laid in the high ground of the south bank of the river, which it crossed near the western limits of the city, and was then connected to a 30-cm, cast-iron pipe in the distribution system. The total length of the pipe from the reservoir to the city distribution system was 2,850 m.

This provisional pipe continued in service from October, 1906, until August 27th, 1909, when the river portion was completely swept away, together with the provisional pump-house at San Geronimo, during the great flood in the Santa Catarina River. Fortunately, the permanent supply works were completed at the time, so that the destruction of this pipe line, which had already served its original purpose, had no effect on the supply of water to the city.

_Infiltration Gallery._--The chief feature of the San Geronimo gravity supply is the infiltration gallery. By referring to the profile on Plate XI it will be seen that at this place there is a considerable area of what is undoubtedly water-bearing gravel. The main conditions were revealed by the borings previously carried across the valley, but the profile has been corrected to show the actual conditions as established at a subsequent date by shafts. Practically, the water-bearing strata are not limited merely to the sand and coarse gravels, as the clay formation lying above and below them is full of small gravel deposits containing considerable volumes of water. The main direction of the underflow is toward the east, and the hydraulic gradient, which was established from wells sunk farther west, was found to be approximately 1%, or practically the same as the average surface of the bed of the river above the line of the infiltration gallery.

The general scheme for tapping this underflow was to drive a main gallery at the 560-m. level on a grade of 0.05%, which was sufficiently high to take the supply by gravity to the western reservoir, having a top water level at 558.75 m. above datum. This elevation is sufficient to give an excellent pressure over about 60% of the city, and a fair pressure to reach the upper stories of the highest houses, if required, over the whole supply district. From this gallery it was proposed to sink shafts at frequent intervals, for a total distance of 300 m., carrying them below the gallery level, to tap any water-bearing gravels there might be in the clay formation underlying the gravels and sands. From the main gallery it was proposed to construct branch galleries up stream on a flat gradient, so as to obtain the advantage of an increased head due to the steep hydraulic gradient of the underflow water.

In investigations of this kind, it is of first importance to have a continuous record of the level of the water plane, and Fig. 12 has been plotted to show its variation at San Geronimo from the beginning of 1905 to March, 1910. From January, 1909, to March 31st, 1910, these levels are averages of daily readings taken in 9 shafts sunk along the proposed line of the infiltration gallery. In 1902 the water plane was standing at 570.18 m. above datum, but from that date until 1905 the writer has been unable to find any records. This diagram should be examined together with the rainfall diagram, Fig. 3, and it will be noticed that the fall in the water plane drops with the general scarcity of the rainfall during 1907-08, and until July, 1909. The year previous to July, 1909, is regarded, by many competent local observers to have been the longest period of extreme drought in 30 years in Nuevo León, and the evidence which the writer has been able to gather regarding stream flow in the neighborhood of Monterrey supports this view. The total rainfall at Monterrey for the year prior to July 1st, 1909, amounted to 9.98 in., or 4.16 in. less than the lowest record for any calender year since 1894, or, in other words, about 45% of the average annual rainfall.