Reservoir management faces a wide range of new challenges resulting from the impact of climate change. One set of challenges arises from the non-stationary nature of hydrological conditions. Another crucial issue is watershed sedimentation, which can significantly influence the sustainability and safety of reservoirs. To address these concerns, this study developed a framework for the management of reservoir risk. An analytical conceptual model coupling physical governing relationships and economic tools was proposed, which was then applied to the Shihmen Reservoir in Taiwan. We adopted a statistical representation of future hydrologic conditions with the assumption of time-variant moments and focused on evaluating the impact of an increase in the frequency of extreme hydrological events caused by climate change and used a stochastic approach to quantify the risk factors. Our results confirm that this approach can be used to identify reservoir-related risks and generate appropriate options for strategy and policy. We determined that the major source of risk is the hydrological conditions, especially the extreme events. More severe intra-annual climatic change is much more dominant in the risk compared to inter-year trends. The influence of reservoir characteristics on risk is associated mainly with the availability of flood control capacity, but limited due to the limitation of its volume and potential to regulate the flow. Engineering may provide an option for mitigating the risk, but integrated, watershed-level approaches, such as providing systematic detention or land use management, are better suited to reducing the storm peak from a long-term perspective. With a critical increase in the risk of overtopping, a high probability of dam failure and corresponding losses may precipitate the need to retire or remove the facility. However, because the benefits and costs are both huge, the decision may be biased by a conservative attitude. The outcome of small facilities failing may be considered more acceptable than similar events besetting larger systems.
Copyright © 2013 Elsevier Ltd. All rights reserved.