FIRMINO, M. B. M.; http://lattes.cnpq.br/2678610871644419; FIRMINO, Mosefran Barbosa Macedo.
Resumo:
The brazilian rivers run across the cities and agricultural and cattle fields while getting
polluted. The modeling and analyses of natural and manmade effects in water quality are,
usually, associated with simulation techniques. However, this technique does not prioritize
optimum water allocation. This paper presents the META-F: Modelo de Enquadramento
Tático das Águas Fluviais (Fluvial Waters Tactical Quality Framing Model) for optimum
water quality and quantity allocation at monthly level, via sequential linear programming,
where the objective function aims the satisfaction of meeting goals to the water quantity
demands and fluvial water quality requirements and the restrictions represent the physical and
operational limitations of the hydraulic components behavior in a water resources system. The
intrinsic non-linearities of each component processes had been implemented through an
iterative procedure, making use of linear approaches.
The model was applied to part of the Capibaribe river stream flow, located between
the Jucazinho and Carpina reservoirs, in Pernambuco state, Brazil, where there are three
irrigated perimeters, two human supply and industrial demands and two reservoirs. The
behavior of the system was analyzed using an average climatic data in three scenarios: for the
average inflow concentration of BOD in two cases, with and without the requirement of
fluvial water quality requirement goal, and, finally, with the maximum inflow concentration
of BOD. The results have shown the model efficiency as well as the viability of applying
linearization techniques to non-separable variables. Still, the difference of attained operation
policies for the reservoirs in the three scenarios can be clearly observed. Moreover, the
robustness of the algorithm can be observed when it a solution for the system is achieved even
when its setup does not generate a variable space that satisfies the water quality requirements