Abstract:
The role of reservoirs to transfer in time water from wet to dry periods is particularly important for water supply systems located in drought-prone areas. Impacts of natural variability of water availability on different water uses can be heavily modified by thorough definition of operating rules of reservoirs. The identification of hedging rules including monthly rationing factors defined for each use could guide water managers on managing foreseen temporary water shortages both reducing their total amount and distribution throughout the year. The aim of this paper is to explore the effectiveness of several types of discrete and continuous hedging rules defined through heuristic optimization and validated by simulation. In particular genetic algorithms have been used to investigate the complex solution domain of rationing factors and storage volumes in reservoir used to trigger rationing on different water uses. The procedure focuses on the definition of the objective function of the optimization problem that plays a crucial role on the achievement of the desired goal. A multiuse water supply system in southern Italy supplying municipal, irrigation, industrial uses and respecting environmental constraints has been used as case study. Implementation of optimized hedging rules improves the performance of the water supply system leading to the reduction of the peaks of shortages and to a better monthly distribution of shortages among uses characterized by different priorities.
