Étude sur modèle réduit de la coupure de la Rance

A model study of the damming of the Rance

Laboratoire National d'Hydraulique, Chatou.

Résumé

1. - Introduction. The considerable tides and currents in the Rance estuary caused a number of delicate problems as regards the construction of the tidal power plant, for as the cofferdamming of the estuary progressed, the water level difference across the structures gradually increased due to the continually decreasing size of the opening available to the tidal flow. One of the most important of these problems, how to ensure the stability of these structures, was studied on an undistorted 1:150 scale model at Saint-Malo near the actual site (fig. 1), which was also used to investigate the various damming schemes for the estuary. II. - Electricité de France scheme. The 800 m long power plant section with thirty-eight 9,000 kW units, ten gated sluiceways and a navigation lock on the left-bank side were to be built on dry land behind three cofferdams, which were to be erected in six phases (fig. 2). The model was used to check the stability of the rock fill platforms, to ascertain the water level differences across the barrage, and to see whether the closure could be carried out according to the proposed programme (fig. 3). III. - Alternatives suggested by various groups of firms. ALTERNATIVE 1 (Fig. 4). This provided for the construction of the power plant section inside a single extensive area protected by prefabricated concrete blacks. The mode tests served to determine the hydraulic conditions to be expected at various stages of work progress and ta design a suitable structural profile. ALTERNATIVE 2 (Fig. 5). This provided for the construction of the power plant section in three separate protected areas. The purpose of the model tests was to investigate the construction of the rock fill dyke and to ascertain the tensile stresses affecting the cables used to tow and anchor the caissons in position. ALTERNATIVE : 1 (Fig. 6). This provided for either two or three cofferdammed areas across the Rance. In the former case, the middle area was to be closed off by a dyke made up of caissons stranded on a rock fill foundation structure, the closure then being completed by filling up the openings between the caissons. The model was used to measure water level differences and currents and to investigate the behaviour of the rock fill foundation structure. In the layout featuring two cofferdammed areas, total closure was to be achieved by means of an articulated caisson, the handling conditions for which were studied on the model. ALTERNATIVE 4 (Fig. 7). This provided for the closure of the estuary by two sites protected by circular gabion cofferdams. The model tests served to measure water level differences under various conditions and to study the stability of the rock fill at the foot of the pile driving platforms. ALTERNATIVE 5 (Fig. 7). In this alternative, the closure was to start with gabions from both banks outwards, the remaning opening then being plugged by tipped rock fill between steel panels. The model tests featured the closure of the final opening in horizontal layers. ALTERNATIVE 6 (Fig. 8). This provided for damming by means of three successive cofferdammed areas and the construction of a submerged dyke with a top rock fill protecting layer before the opening-up of the left-bank area. This was to be followed either by tipping rock fill to progressively complete the closure, or by stranding caissons for the porpose.The possibility of replacing the dyke by four big caissons was also considered. The model tests served to check rock fill stability and how easily the caissons could be brought into position. ALTERNATIVE 7 (Fig. 9). This provided for the construction of part of the power plant section on a site protected by an initial gabion cofferdam, and further construction on a second similar site protected by a preliminary closure dyke. Currents and loads on the outermost gabions were determined at various conditions on the model. ALTERNATIVE 8 (Fig. 9). This scheme featured three areas protected by gabion cofferdams, the problem being to find out whether it would be worth while completing the right-bank cofferdams in the shelter of a rock fill dyke. The model was used to determine currents, loads acting on the gabions and conditions for the construction of the dyke. ALTERNATIVE 9 (Fig. 10). Successive stages provided for in this scheme were : 1) construction of a gabion cofferda on the right bank side, 2) construction of a gabion dyke, and 3) construction of a rock fill dyke to complete the closure. The model was used to determine thrust forces on the gabions, how far to build the initial right-bank gabion structure out into the river, also currents and flow reversal duration near the structures. ALTERNATIVE 10 (Fig. 10). This provided for a lock construction site on the left bank side, a cofferdam at the + 14 m level cutting the estuary off from the sea, and the construction of the dyke on the south side at the + 9 m level. The interesting part of this scheme is the north dyke formed of prefabricated reinforced concrete caissons with sheet piling gabions between them. The plan was to first sink the reinforced concrete caissons so as to partly block the estuary and then to build the interconnecting gabions so as to form a crenelated dyke, after which the final closure would be achieved by installing the sluiceways. Conditions for the stranding of the caissons were investigated on the model. ALTERNATIVE 11 (Fig. 11). In this alternative, the lock cofferdam was to be erected first, followed by the initial cofferdam structures. With the latter under construction and the caisson stranding operations in progress, the rock between the two initial cofferdam structures was to be levelled off with the aid of a caisson, and a concrete sill built in the estuary channel up to the - 10 m level. Twenty-eight 9m diameter reinforced concrete cylindrical caissons were then to be stranded in the positions prepared with the aid of the rock-levelling caisson, the gaps between them being plugged at slack water by reinforced concrete planks. This structure was to enable sheet piling cells to be erected in a position sheltered from the flow. The model was used to determine how far to extend the initial gabion cofferdam structures, slack water duration, water level differences across the dam, and flow velocities in the gaps between the caissons. IV. - The 1960 project (Fig. 12). This has been described in a paper by Mr. Duhoux. The model served an important purpose in the designing of this project by enabling the following to be determined : (i) A construction programme for the gated sluiceway cofferdam. (ii) How far out to build the initial gabion structures for the main cofferdammed area. (iii) A programme for the positioning of the reinforced concrete caissons and the construction of the cells. (iv) Stresses likely to affect the cofferdam structures due to waves propagating in the estuary. It is very interesting to compare observations on site with the corresponding model data. This point is also discussed in Mr. Duhoux's paper.