Author(s): S. Hatchard; R. J. P. Schmitt; F. Pianosi; J. Savage; P. D. Bates
Keywords: Hydropower optimisation; Reservoirs networks; Downstream impacts of dams
Abstract: Development of hydropower in developing countries carries economic development rewards, and significant environmental impacts, due to fragmentation in river networks . A key environmental impact associated with storage hydropower is the reduction of lateral floodplain connectivity, which results in reduced floodplain inundation, which negatively impacts biodiversity . Determining the optimal site, design, and operation (SDO) of dam portfolios which maximises power generation and minimises this environmental impact is a complex problem. Simulation - Optimisation models of hydropower portfolios have often included impact on annual flood peak as a proxy objective for floodplain impacts, but have rarely explicitly included inundated area as an objective. Furthermore, when done this type of analysis is done, it is usually performed at a monthly timescale, which underestimates flood peaks and neglects in-channel and floodplain hydraulics. This work presents a multi-dam simulation - optimisation framework which uses the outputs of high-resolution hydrodynamic modelling to explicitly model the impact of different portfolio’s SDO on inundated floodplain extent, and to include this modelled extent as an objective. This incorporates channel and floodplain hydraulics at a fine time resolution, allowing a more realistic representation of the impact of hydropower development on biodiversity, setting it as an objective in the optimisation framework. The optimisation framework is applied to the data scarce Pungwe Basin in Mozambique / Zimbabwe, and identifies significant trade-offs from mainstem damming between power production and downstream hydraulic connectivity between rivers and floodplains. The inclusion of hydraulically modelled inundated area represents a step forward for increasing the ability of simulation - optimisation frameworks to model the real consequences of hydropower development, and thus can allow real biodiversity constraints to be set as modelling objectives.