Skip to Main Content U.S. Department of Energy

Hydro Management

Next-generation tools for improving energy system operation efficiency

Modern society depends on a reliable supply of water and energy. Critical infrastructure including dams, canals, energy plants and energy transmission systems have been developed at considerable cost to ensure the public has adequate and reliable sources of water and energy. The continuing increase in demand for water and energy combined with the limitations of existing infrastructure capacity have required balancing capacity expansion with improved operation efficiency through more intelligent real-time operation.

Managing energy and water resources requires working within the margins — looking for increased operational efficiency, rather than wholesale changes to the infrastructure. Improved methods for stream flow forecasting and new tools to access and manage uncertainty will provide resource managers the opportunity to better optimize their operations. More then 80 percent of the water supply in the Western United States is provided by snowmelt runoff. This supply is variable and uncertain, while storage is diverse and fragmented. Additionally, considerable uncertainty exists in the forecasts of future meteorological conditions, which are the primary drivers of the hydrologic processes.

Water resources management system contributes to operation efficiency

Multiscale hydrological modeling
Figure 1. Multiscale hydrological modeling showing a) HUC6 representation of the Columbia River Basin using over 35,000 sub-basin/elevation bands at an average spatial scale of ~16 km2; b) DHSVM simulated snow water equivalent in the 1,600 km2 Big Wood Basin at a 180-m (0.03 km2 ) spatial resolution. Click for a larger image.

Example of methodology for ensemble stream flow forecast
Figure 2. Example of methodology for ensemble stream flow forecast in the Yakima River Basin beginning in a) November and b) December. The single blue line represents observed stream flow up to the start of a forecast period. Gray lines represent ensembles of stream flow forecasts using physical process models driven by ensembles of possible future meteorological realizations. Note how the forecast uncertainty (particularly for May/June) decreases between the forecasts made in November (top image) and December (bottom image) as additional observations are included. Click for a larger image

At Pacific Northwest National Laboratory, we are developing next-generation tools in the form of a scaleable water resources management system to assist Pacific Northwest National Laboratory's Electricity Integrated Operations Center (EIOC) improve energy system operation efficiency. This system will use improved snow/water forecasts to clearly articulate tradeoffs between multiple energy-related water management objectives. It is designed to take full advantage of the computational assets, data processing, advanced visualization capabilities, and real-time data feeds supplied by the EIOC to provide data critical to more efficient operation of a constrained energy system.

We began by focusing on water budget characterization and improved snowmelt forecasting. This required development of a regional information infrastructure to support multiscale hydrological modeling and analysis (Figure 1) to better understand and ultimately predict the location, capacity, storage and fluxes of water in the Columbia River Basin. Computational methods have been developed to extend the snowmelt simulation approach and climate forecast models to develop daily ensemble stream flow forecasts. Uncertainty in forecasts is addressed explicitly (Figure 2) during generation of ensembles of stream flow forecasts using physical process models that are driven by ensembles of possible future meteorological realizations. This approach allows the model to be updated with EIOC-supplied real-time data feeds of spatial snow cover and stream flow using an Ensemble Kalman (EnKF) data assimilation strategy. The EnKF accounts for uncertainty in weather forecasts, model parameters and observations used for updating. Making the ensemble generation system capable of supporting real-time management decisions requires computational assets such as those provided by the EIOC.

A multiscale, multiple objective optimization system is also being developed to demonstrate how large-scale optimization methods can be used to control complex water and energy systems at the granularity of a single turbine to a system as large as the Pacific Northwest energy and water system. This effort will demonstrate optimization of a hydropower reservoir system linked with a wind system. Both hydropower and wind systems will rely on weather-based ensembles at a variety of temporal scales. The hydrosystem will be managed to compensate for the intermittency of the wind system. Efforts have also begun to integrate habitat constraints such as anadramous fish migration into the water resources management system.

Publications and Presentations

Cook CB, R Skaggs, LW Vail, and MS Wigmosta. 2006. "Water Resources Management within the Electricity Infrastructure Operations Center." Presented by Christopher Cook at Internal Student Environmental Development Program (SEDP), Richland, WA on August 24, 2006. [Unpublished]

About the EIOC




Additional Information

Research Areas

Deploying Technologies

Website Contact