A Fuzzy Multicriteria Approach to Estimate the Water Quality Index of Hydroelectric Reservoirs

The issue of water contamination in reservoirs and watersheds due to the irregular use of fertilizers and uncontrolled occupation by riverside populations has worried managers operating hydroelectric plants. The increase in the levels of certain physical–chemical parameters causes problems of accelerated deterioration of equipment and clogging of water pipes. Within this context, here we estimate a specific water quality index (WQI) for hydroelectric plants. The novelty of this article is to use expert consultation to select the most critical parameters that cause damage to the dam structures and mechanical equipment. We propose a method carried out in three steps: selection of specific parameters for hydropower plants and fuzzy MCDA estimation of their weights; fuzzification of sub-indices; and estimation and classification of the WQI. In this last step, the sub-indices are defuzzified and weighted for classification according to the degree of membership in a fuzzified scale. Fuzzy logic deals with variables that do not allow objective classification, given that the limits between classes are imprecise or non-statistical, such as the parameters analyzed in this study. These thresholds are fuzzy numbers, so the treatment of the sub-indices must follow this logic. We obtained a specific classification for the water quality upstream of the Corumbá IV Hydroelectric Plant, located in the central region of Brazil. Observing the individual ratings of each parameter obtained before aggregation, although producing results for drinking water set by Brazil’s National Environmental Council, presents concentrations of some parameters that, if not controlled, could harm the hydroelectric plant.