GOMEZ Eduardo Alberto
congresos y reuniones científicas
Hydrological Modeling and Management of a Salt Lake with Dynamic Optimization Strategies.
Congreso; World Congress of Chemical Engineering; 2017
Institución organizadora:
Permanent salt lakes are continuously threatened by changing hydrological and salinity regimes, which may be in turn caused by extended rainy or drought periods and water diversions in the basin. These ecosystems present research challenges regarding conservation and management issues. The salt lake Chasico, in Argentina, is a water body in an endorheic basin, with semiarid climate. It has undergone extreme ecologic events during the last century, which include severe droughts and floods, being the largest one in 1983. The lake vulnerability to meteorological changes affects human, agricultural and touristic activities, as well as sports fishing of silverside (Odonthestes bonariensis). In this work, we formulate a hydrological model within a dynamic optimization framework for lake Chasicó. The mathematical model comprises a set of differential algebraic equations that stand for mass balances for water and salt, as well as evaporation, calculated through energy and momentum balances; bathymetric information to assess lake volume at two points in the time horizon, etc. The time unit is one day. Forcing functions include tributary flow rate profiles, underground water flow profiles, temperature, relative humidity, solar radiation, mean wind rate and related meteorological profiles along a time horizon of eight years and precipitation profiles. Additional algebraic equations include latent heat of vaporization and vapor pressure correlations for water, among others. The lake volume decreases from an initial value of 475 hm3 to 273 hm3 (2008-2013), with a consequent increase in salinity from 23 to 41 g/l. During the last 3 years, volume increases and salinity reaches 32 g/l. We validated the model with collected salinity data throughout this period. As a second step, we studied two opposite potential scenarios. The first one considers a drought period (accumulated mean annual precipitation of 465 mm), starting at salinity values of 41 g/l. At the end of a time horizon of 8 years, final salinity value is 97.6 g/l. The second scenario addresses a rainy period, which begins at current salinity conditions and reaches salinity values of 16 g/l in 8 years. We formulated an optimal control problem to manage the lake in rainy periods; the objective is to avoid flooding of the nearby village and tourist facilities, reduce investment risk, as well as to maintain salinity at appropriate levels for silverside reproduction and sustainable fishing. The control variable is a fraction of total tributary flowrate that would be derived through an alleviation channel. The performance criterion is the minimization of an integral objective function (the square difference between actual salinity and a desired value), subject to the differential algebraic model and additional path constraints. In this way, the model provides a decision-making tool to plan management actions, as well as to estimate the constructed channel size and cost.