عنوان مقاله [English]
Recently, flood crisis due to the extreme rainfall events and climate changes are increased. There are several structural and non-structural methods for flood control and mitigation. The detention dams are structures used to attenuate the flood peak and reduce flood damages. The detention rockfill dams constructed by natural rockfill with porous media have interesting applicability for flood mitigation. The design approach for the detention rockfill dams include three main steps. 1- In the first step the location of the dam is done to regulate the reservoir volume required to decrees the flood peak with specified return period in such a way that provide the safe deplete flood condition in the downstream.; 2-In the second step the preliminary designs are done to design the detention rockfill dam and derive the preliminary dimensions and sizing the dam. 3-In the third stage the optimum dimension of the dam is determined using the preliminary designs done and single or multi objective optimizations. Based on these steps, there are several graphical or mathematical methods for preliminary designing of detention rockfill dams. The graphical methods are developed and used in early studies (Akan, 1990; McEnroe, 1992; Abet & Grigg, 1978; Wycoff & Singh, 1976). In these studies, the designs are done using single or double output pipes, without any infiltration form the dam body. In these studies, the depth-storage relation of dam is linear assumed while in the detention rockfill dams this relation is nonlinear. The graphical method of SCS uses two curves of peak storage ration to total flood volume and ration of peak flood discharge to beak inlet discharge. Also, in some studies simple design equations are developed.
In this paper, using a simulator-optimizer model, the detention rockfill dam is optimized based on reliability. The case study of the paper is the design of detention rockfill dams for flood control in Karganroud river in the west of Gilan province based on hydrometry data 1957-2017, which can reduce the peak of flood based on preliminary design, single optimization, and reliability-based design optimization. The model is a multi-level model. The hydrological equation of nonlinear flood routing with non-Darcian flow inside the rockfills is used for preliminary designs. The structural stability analyses is used for the safety factor calculation and reliability determination. The optimization techniques is sued to derive the optimal design regarding the stability and reliability analysis. Using the governing equations, the dam can be simulated, and then the preliminary design can be gained for four different angles of the dam wall which are equal to 90, 80,75, and 70 degrees. Then these preliminary designs are redesigned to satisfy the reliability constraints in the optimization module. Based on the preliminary designs, the optimum design derived using Monte-Carlo simulation for uncertainty and Genetic Algorithm for multi objective self-adaptive optimization.
Based on the results, the stability criteria in preliminary designs are not satisfied based on the calculated SFS, SFO and SFF factors, so these dimensions are used as lowest limits in the optimal designs. The uncertainty analysis in the input parameters and the hydraulic performance of the dam under uncertainty is determined and reliable designs are selected. of the model is investigated using Monte-Carlo simulation The output results of the multi-objective algorithm have a wide range of different kinds of designs that can be made a decision based on them. The output results of single-objective and multi-objective optimization show that the optimal situation between four different situations occurred when the angle of the dam wall is equal to 80 degrees. The pareto front of costs versus reliability provide the final design of the dam regrading selected reliability. In this situation, the cost of construction is reduced by 20%. Also, the results confirmed that with reducing the angle of slopes, the construction cost reduced round 40%. The developed model in the current study by using uncertainty, optimization and reliability is capable in redesigning of detention rockfill dams in applied flood control plants.