Modifications hydrologiques d'un cours d'eau péruvien après mise en eau d'un barrage-réservoir

Hydrological alterations of a peruvian river downstream of a new impouding reservoir

Degrémont 183, avenue du 18-Juin-1940, 92508 Rueil-Malmaison Cedex - France

Résumé

La mise en eau d'une retenue au Nord du Pérou a été suivie d'une période exceptionnelle de sécheresse étalée sur cinq ans. Le débit réservé prévu n'a pu être respecté, ce qui a permis la fixation d'îlots sableux d'origine éolienne dans le lit de la rivière en aval du barrage ; en outre, ce sable est fortement imprégné de chlorures et de sulfates. Les conséquences de cette situation sont multiples: limnologiques (augmentation excessive de la salinité de l'eau), hydrologiques (inondations lors du retour des pluies) et technologiques (corrosions du matériel métallique de traitement et de distribution de l'eau potable); l'étude et l'analyse de ces phénomènes font ressortir la difficulté de prévoir tous les aspects de l'impact que la construction d'un nouveau barrage peut exercer sur le milieu naturel.

Abstract

The impoundment of the Poechos dam (maximum capacity: 1000 Mm³), on the Rio Chira in northern Peru, coincided with the beginning of a drought which lasted for several years: the total annual inflow was about 2000 Mm³ from 1978 to 1982, in contrast to the typical 5000 Mm³.

This coincidence led to very low discharge downstream, even during the rainy season. In contrast to the previous usual values of several hundred, or even several thousand m³/s, the peak flows dropped as low as around 20 m³/s from 1978. Likewise, dry weather flows fell from 10-20 m³/s to 3-5 m³/s over the same period.

This coastal region is notable for the wind that constantly pushes sand dunes (barchans) along the surface of the earth and dumps the sand into the river bed. These quantities of sand impregnated with chlorides and sulfates were formerly carried off by the floods. When the floods ceased due to the drought and the filling of the dam, the sand masses remained in the river bed, particularly as the spontaneous growth of shrubbery helped to hold them in place. The simultaneous presence of the sand banks and the low river discharge led to a sharp rise in the chloride and sulfate contents of the river water.

This high chloride level caused corrosion of the metal equipment of the treatment plant and distribution system. A nearly one-year study conducted on control specimens made of ordinary black steel was undertaken to prevent or at least limit the corrosion. The return of the floods while the study was in progress, at the same time as it caused disastrous inundations of the whole valley, permitted measurements of the rates of corrosion resulting from waters of widely different characteristics (particularly in terms of alkalinity and chloride content); the results of this experimental study have been published in a previous paper (Mouchet, 1997).

These observations lead to the conclusion that even after comprehensive consideration of the consequences of river impoundment within a long-term perspective, problems may arise from a lack of complete understanding of the dynamic nature of the basin.