Minimization of the hydraulic impact of a physical connection through a river flood plain

A physical connection that extends through the flood plain of a river produces a perturbation of the flow during floods. Projects of this type are presently the focus of environmental concern. They should guarantee that the perturbation on the flow is minimal, so as to avoid a situation which could lead to geomorphologic changes incompatible with the natural evolution of the system. To accomplish this objective, a thorough study of the fluvial system is required, in order to, in the first place, know the conditions before the construction of the connection and, in the second place, predict its response in the future situation, once the works are done. This type of study must proceed through the following steps: 1) Characterization of the pre-operational stage of the system; 2) Implementation of hydrodynamic mathematical models of the river flow, calibrated so as to simulate accurately recorded pre-operational flood situations; 3) Exploitation of the models so as to predict the future behavior of the system; 4) Definition and application of a criterion of ‘acceptable impact’. The detailed methodology of such a study is illustrated through its application to a particular project: the physical connection between the cities of Rosario and Victoria (Argentina), which proceeds through the Paraná River valley, of about 60 km width at that location. The Paraná River is one of the major watercourses of the world (mean discharge is 18,000 m 3 /s, with recorded peaks up to 60,000 m 3 /s). The flood plain at the project location is traversed by a large amount of permanent and non permanent secondary branches. The connection is presently under construction by the concessionaire Puentes del Litoral S.A., a consortium formed by Impregilo-Iglys, Hochtief, Benito Roggio e hijos, Sideco-Iecsa and Techint, and the hydraulic project and environmental impact study were done by Serman & Associates Consultants. The pre-operational stage of the system was characterized through a database containing: a) Hydrographic data: Records and statistics; b) Topographic data: Historical data complemented with an ad-hoc survey; c) Streamlines data: Through interpretation of satellite images for one of the major floods (1998); d) Geomorphologic data: Through interpretation of satellite images corresponding to low water conditions; e) Vegetation data: Using a different processing for the low water satellite images. Two hydrodynamic models were implemented: a one-dimensional channel network model for the regional scale and a two-dimensional model for the local scale. The branches of the 1D model were built based on the streamlines data. This model extended from 80 km upstream (Diamante city) to about 300 km downstream of the connection. At the upstreamost section (Diamante city) the time series for the discharge was imposed as a boundary condition, while the downstream end of the model (‘Rio de la Plata’) was a tidal boundary. This model provided the boundary conditions for the 2D model, which included all the valley width and the main branch of the river, from about 30 km upstream to 30 km downstream of the connection. For both models, the quantification of the hydraulic resistance, through a roughness coefficient, was made based on an analysis of the geomorphologic and vegetation zones. The models were calibrated for two different situations: extraordinary floods with normal duration, using data from the recent 1998 flood, and extraordinary floods with abnormally long duration, which only record is the 1983 flood. The main difference between them was the vegetation collapse during the long duration flood, which led to a decreased hydraulic resistance. The models were then used to simulate friture scenarios with different alternatives for the bridges widths and locations, for a total transparency for the flood plain not lower than 18% (a percentage that must guarantee a water level increase upstream of the connection lower than 40 cm for the design flood, according to previous preliminary studies).