A SIMULATION - OPTIMIZATION APPROACH IN DEVELOPMENT OF OPERATION POLICY FOR A MULTIPURPOSE RESERVOIR

摘要

In the present study, behaviour of a reservoir system, Ukai reservoir, India has been examined by using a simulation model with reservoir sedimentation and flood control operating scenarios. A standard operating policy (SOP) based simulation model has been developed considering historical reservoir data of 36 years from 1975 - 2010. The system behaviour has been assessed using statistical performance indices like reliability, resilience and vulnerability. The simulation results indicated that the system is highly reliable for water supplies on one hand, and vulnerable to flood damages on the other. Keeping in view the need to optimize the system with the objective of minimizing the flood damages, ensuring reliable water supply for municipal and industrial, and irrigation demands, a monthly time stepped backward recursive stochastic dynamic programming (SDP) model has been developed while giving due consideration to target releases and target storage for monsoon periods with appropriate flood control constraints. In the SDP model, initial reservoir storage and inflow, in a particular month, have been treated as state variables; and stochasticity of inflows are addressed by computing one-time-step transition probabilities. The reservoir storage and inflow each have been discretized into seven class intervals. The optimal reservoir releases and optimal end of the period storages are reported for different combinations of state variables. The optimal policies are synthesized and rule curves are proposed for the optimal operation of Ukai reservoir. The simulation-optimization approach indicates an improvement in the reservoir operation in terms of increasing storage volumes, ensuring reliable releases, and decrease in spill events compared to existing operating conditions.