Tradeoffs associated with sediment‐maintenance flushing flows: a simulation approach to exploring non‐inferior options

Abstract

AbstractSediment‐maintenance flushing flows designed to mimic the action of natural floods in removing the accumulated fine sediments from the channel and loosening the gravel bed have been increasingly proposed as an effective alternative in dam management and a required component of riverine restoration programmes. However, reservoir releases are generally associated with financial and environmental costs, thus it is highly desirable to specify flushing flows as accurately as possible. In this paper we present a simulation approach to evaluating flushing flows and exploring the tradeoffs associated with non‐inferior flushing options. A two‐fraction sediment routing model is used to simulate the gravel‐sand bed response to flushing flows. The results reveal that the sand cleansing effect propagates from upstream to downstream and from surface to subsurface. Under a steady gravel supply from upstream, an equilibrium state of gravel transport and bed degradation is eventually reached in the simulation reach. The flushing flow and sediment transport system investigated in this study involves a transient state variable (bed sand content), a decision variable (flushing flow discharge), a flushing goal (ultimate bed sand content), and three outcomes to be minimized (flushing duration, released water volume, and total gravel loss). A series of numerical simulations are carried out with a range of flows and pre‐flushing bed sediment conditions. The results reveal that the flushing efficiency is higher for the larger flow. However, for flows greater than ∼100 m3/s the flushing duration is less sensitive to the flow discharge, thus the system may be simplified as a bi‐objective one. The gravel loss and water volume are two conflicting outcomes within the non‐inferior flow region. Under a worse bed sediment condition, the feasible flushing options are constrained in a narrower range and also associated with higher costs. The tradeoffs between the conflicting outcomes are quantitatively displayed with the transformed feasible solutions in the objective space. We provide here a general and practical approach permitting a quantitative evaluation of the different flushing options that is appropriate to the level of data typically available. Copyright © 2004 John Wiley & Sons, Ltd.