Optimization of Multi-Reservoir Water Supply Operation Utilizing Ensemble Inflow Forecasts

The Dallas-Fort Worth Metroplex is one of the fastest growing urban areas in the US. The Tarrant Regional Water District (TWRD), one of the largest raw water suppliers in the region, operates a system of reservoirs to meet the growing demand, including moving water from the wetter east to the drier west which incurs significant energy cost. Recently, a set of ensemble hindcasting experiments were carried out to assess the value of ensemble inflow forecasts generated with the NWS’s Hydrologic Ensemble Forecast Service (HEFS) in TRWD’s water supply system operation. It was found that one month or less-ahead HEFS inflow forecasts based on the Global Ensemble Forecast System (GEFS) have significant skill, but that the impact to the system operation under the existing operating policies and constraints is very limited. This research aims at identifying the limiting factors, optimizing the operating policies under the current constraints, and exploring opportunities for improving cost effectiveness and reliability under varying levels of stringency in the constraints. In this presentation, we describe the ensemble prediction-optimization framework and share preliminary results and findings.