Optimization of Multi-Reservoir Management Rules Subject to Climate and Demand Change in the Potomac River Basin

Abstract

Water management in the Washington Metropolitan Area (WMA) is challenging because the system relies on flow in the Potomac river, which is largely uncontrolled and augmented by the Jennings-Randolph reservoir, located 9-10 days travel time upstream. Given this lag, release decisions must be made collectively by federal, state and local stakeholders amid significant uncertainty, well in advance of accurate weather forecasts with no ability to recapture excess releases. Adding to this uncertainty are predictions of more severe and sporadic rainfall over the next century, caused by anthropogenic climate change. This study aims to evaluate the potential impacts of demand and climate change on the WMA water supply system, identifying changes in system vulnerability over the next century and developing adaptation strategies designed to maximize efficiency in a nonstationary system. A daily stochastic streamflow generation model is presented, which succesfully replicates statistics of the historical streamflow record and can produce climate-adjusted daily time-series. Using these time series, a multi-objective evolutionary algorithm is used to optimize the system's operating rules given current and future conditions, considering several competing objectives.