Metodologia multi-objetivo para alocação da vazão excedente em bacias hidrográficas.

Abstract

This work assumes that, in Brazil, water rights concession criteria are very restrictive, generating conflicts among users. In addition, the climate variability causes changes in water availability, so that in periods of higher flows, there is an excess discharge that is unavailable for the production of social wealth. This Thesis argues that the discharge surplus could be allocated among users, then minimizing losses due to not using it; the allocation can be defined annually, so that to consider the inter-annual variability of the hydrological variables. The allocation of a variable amount of water availability provides a better utilization of water resources, and is an adaptation strategy to cope with climate variability. However, it is true that the development of a decision support system for allocating the surplus discharge is not simple, because it is necessary the involvement and integration of many variables and methods that must be integrated in an appropriate optimization model. The approach should also be multiobjective and multicriteria and should be integrated with the water resources management law, in order to fill the gaps without violating the existing rules. Thus a methodology is proposed for optimizing the allocation of surplus discharge in a basin through a multiobjective evolutionary algorithm in which adaptations and reproduction operators were inserted to incorporate the specific constraints of the problem and overcome the obstacles presented. In addition this evolutionary algorithm was developed in order to be easily coupled to other models and be easily adaptable to scenarios of rationing (zero or negative surplus discharge). The methodology was evaluated in the Gramame river basin, in the state of Paraiba, which is already presenting evidence of greater demands than availability. The analysis included two hypothetical scenarios of probabilistic forecasts of precipitation: one above-normal forecast and one below-normal forecast; they allowed the assessment of the behavior of the optimization model in situations of allocating the surplus discharge as well as rationing when necessary. We also analyzed the behavior of the model with the adoption of management strategies with the flexibility of environmental flow values and allowable BOD concentration. The robustness of the allocation strategies were evaluated against changes in historical hydrological variables, which may be caused by climate change or even be the result of uncertainties. The results show that the algorithm proved to be adequate, presenting convergence for the most viable regions of Pareto Front.